Вы находитесь на странице: 1из 159

A Guide to

DP Electrical Power and


Control Systems

IMCA M 206
November 2010

International Marine
Contractors Association
www.imca-int.com
AB








AB
The International Marine Contractors Association
(IMCA) is the international trade association
representing offshore, marine and underwater
engineering companies.

IMCA promotes improvements in quality, health, safety,
environmental and technical standards through the publication
of information notes, codes of practice and by other
appropriate means.

Members are self-regulating through the adoption of IMCA
guidelines as appropriate. They commit to act as responsible
members by following relevant guidelines and being willing to be
audited against compliance with them by their clients.

There are two core activities that relate to all members:
Competence & Training
Safety, Environment & Legislation

The Association is organised through four distinct divisions,
each covering a specific area of members interests: Diving,
Marine, Offshore Survey, Remote Systems & ROV.

There are also five regional sections which facilitate work on
issues affecting members in their local geographic area
Asia-Pacific, Central & North America, Europe & Africa, Middle
East & India and South America.


IMCA M 206


This report was prepared for IMCA, under the direction of its
Marine Division Management Committee, by GL Noble Denton.


www.imca-int.com/marine



The information contained herein is given for guidance only and endeavours to
reflect best industry practice. For the avoidance of doubt no legal liability shall
attach to any guidance and/or recommendation and/or statement herein contained.
2010 IMCA International Marine Contractors Association

A Guide to DP Electrical Power and Control Systems

IMCA M 206 November 2010



1 Introduction ........................................................................................................... 1
1.1 General Introduction and Scope of Document .............................................................................................. 1
1.2 Electrical Propulsion .............................................................................................................................................. 1
1.3 Power Station Concept ........................................................................................................................................ 2
1.4 Conventional Propulsion ...................................................................................................................................... 2
1.5 Concepts of Fault Tolerance and Control Systems ...................................................................................... 2
1.6 DP Class 2 and DP Class 3 Redundancy Concepts ....................................................................................... 4
1.7 Failure Modes and Effects ..................................................................................................................................... 4
2 Power Generation ................................................................................................. 7
2.1 General Principles of Power Generation ......................................................................................................... 7
2.2 Engines.................................................................................................................................................................... 16
2.3 Engine Auxiliary Systems ................................................................................................................................... 22
2.4 Engine Control and Safety Systems ................................................................................................................ 27
2.5 Safety Functions ................................................................................................................................................... 30
2.6 Generators ............................................................................................................................................................ 32
2.7 Fuel Control ......................................................................................................................................................... 34
2.8 Excitation Control .............................................................................................................................................. 43
2.9 Main Switchboards and Motor Control Centres ........................................................................................ 47
2.10 Power System Faults........................................................................................................................................... 48
2.11 Overall Protection Philosophy ......................................................................................................................... 49
2.12 Generator Protection Philosophy ................................................................................................................... 59
2.13 Advanced Generator Protection (AGP)........................................................................................................ 63
3 Power Management ............................................................................................ 66
3.1 Requirement for a Power Management System .......................................................................................... 66
3.2 Power Management System Architecture .................................................................................................... 66
3.3 Power Management System Hardware ......................................................................................................... 67
3.4 Power Management Functions ......................................................................................................................... 68
4 Power Distribution .............................................................................................. 82
4.1 Power Distribution Schemes ............................................................................................................................ 82
4.2 Power Plant Configurations .............................................................................................................................. 90
4.3 Operational Configuration .............................................................................................................................. 102
4.4 Transferable and Dual Fed Consumers ....................................................................................................... 103
4.5 Battery Systems ................................................................................................................................................. 104
4.6 Separation of Redundant Elements for Fire and Flood ............................................................................ 105
5 Thrusters, Drives and Controls........................................................................ 107
5.1 General Propulsion Principles ........................................................................................................................ 107
5.2 Thrusters ............................................................................................................................................................. 108

5.3 Thruster Auxiliary Systems ............................................................................................................................. 115
5.4 Thruster Motors ................................................................................................................................................ 118
5.5 Variable Speed Drives ...................................................................................................................................... 119
5.6 Thruster Control Systems .............................................................................................................................. 122
6 Safety Systems ................................................................................................... 126
6.1 General Principles of Safety Systems ............................................................................................................ 126
6.2 Regulations Relating to Shutdown Systems ................................................................................................ 126
6.3 ESD Systems and DP Redundancy ................................................................................................................ 127
6.4 Active Fire Protection ...................................................................................................................................... 129
6.5 Effects of Ventilation System Shutdown ...................................................................................................... 131
6.6 Group Stops ....................................................................................................................................................... 132
6.7 Control Room Layouts .................................................................................................................................... 132
7 Vessel Management Systems ........................................................................... 134
7.1 General Description ......................................................................................................................................... 134
7.2 Network Technology ....................................................................................................................................... 139
7.3 Redundancy ......................................................................................................................................................... 147


Appendix
1 Abbreviations List .............................................................................................. 151

IMCA M 206 1
1 Introduction
1.1 General Introduction and Scope of Document
This handbook is intended to provide a reference guide for those involved in all aspects of dynamic
positioning (DP), with particular relevance to those involved in designing, assessing and maintaining
dynamically positioned diesel electric vessels.
This document consolidates and updates three DPVOA/IMCA guidance documents:
126 DPVOA Reliability of electrical systems on DP vessels;
108 DPVOA Power system protection for DP vessels;
IMCA M 154 Power management system study.
The previous documents were closely related and published between 1992 and 2000. Since there was
a need to incorporate technological advances it was considered appropriate to assemble all of the
information in one comprehensive document. This guide discusses the DP system, focusing on the
electrical components, and covers:
power generation;
power management;
power distribution;
vessel management systems;
thruster drives and controls;
safety systems.
Under each topic the guide explores early systems and the latest technology available. Some of the
earlier systems are still on older vessels and their technology remains relevant. The operating
principles, advantages, limitations and fault patterns of the components and systems are discussed with
particular reference to FMEAs, trials and to satisfying classification society requirements for
redundancy and recovery from faults. The advantages and limitations impact the complete vessel
design in terms of, for example, cost to purchase, cost to operate, space requirements, operability,
operating procedures etc. Therefore, there is information in this guide which is relevant to a wide
variety of users including:
vessel owners and operators;
vessel designers;
equipment manufacturers and suppliers;
inspectors auditors and consultants;
class;
clients, etc.
For each of those users, there will be information in this guide that is useful to managers, marine
departments, vessel crew, procurement personnel, and frontline engineers. This guide contains aides-
mmoire, recommendations for studies and further analysis, comparisons and other material which will
help users to deliver their work better towards building, running and maintaining safe and efficient DP
vessels.
1.2 Electrical Propulsion
Electric propulsion using steam turbines or diesel engines as the prime mover has been used in ships
since the early 1900s. Diesel electric propulsion is now almost universal amongst medium and large
DP vessels but direct driven and hydraulically driven thrusters are still used in certain applications.
In its simplest form, diesel electric propulsion consists of a diesel engine driving an electrical generator
which is connected to a motor driven propeller or thruster by way of an electric cable.

2 IMCA M 206
The generator and motor may be of the alternating current or direct current type, or of different
types if converters are included. The thrust developed by the propeller may be controlled by varying
the speed of the engine, the speed of the motor, or by varying the pitch of the propeller. Such simple
systems can still be found in certain applications but the vast majority of medium to large DP vessels
have a diesel electric power plant based on the power station concept.
1.3 Power Station Concept
In the power station concept, electric power is provided by several synchronous alternating current
generators operating in parallel. The generators are connected to switchboards by way of circuit
breakers that allow the generators and loads such as thrusters, service transformers and motors to be
connected and disconnected as required. Typical power plants have four, six or eight generators
connected to two or more switchboards.
The advantages of the power station concept include:
greater freedom in location of engines and thrusters;
ability to provide large amounts of power for activities other than propulsion;
ease with which power can be distributed for auxiliary systems;
ease of thrust control for systems of multiple propellers;
well suited to fault tolerant and redundant propulsion systems;
modular designs allow maintenance to continue during operations;
flexibility in engine assignment;
good power plant efficiency.
The disadvantages of the power station concept include:
high initial cost;
complexity;
specialist maintenance requirements;
the efficiency of the power plant can be reduced in designs where fault tolerance is dependent on
maintaining a large spinning reserve. This can also introduce maintenance and emission control
issues. This problem is not unique to diesel electric plant and effective load shedding measures
can improve the situation.
1.4 Conventional Propulsion
Conventional propulsion systems are still widely used in DP applications usually in combination with
limited diesel electric systems. Offshore supply vessels and anchor handlers will often have
conventional marine engine installations with gearbox, shaft line and controllable pitch propellers to
provide high transit speeds and along ships thrust. A combination of shaft alternators driven by the
main engines and auxiliary generators provide power for thrusters and auxiliary systems.
Some offshore supply vessels use direct diesel driven thrusters. Thrust is controlled by varying the
engine speed over a certain range and by varying pitch at lower speeds. In some arrangements a
slipping clutch is used to control speed at the lower speed range when a fixed pitch propeller is used.
Thrusters driven by a hydraulic motor and power pack are also used in some applications.
1.5 Concepts of Fault Tolerance and Control Systems
All DP Class 2 and DP Class 3 vessels are required to be fault tolerant. Although there are slight
variations between the major classification societies, almost all have requirements similar to those
described in IMO MSC 645 Guidelines for Vessels with Dynamic Positioning Systems, which states in
Section 2.2:
For equipment Class 1, loss of position may occur in the event of a single fault;

IMCA M 206 3
For equipment Class 2, loss of position is not to occur in the event of a single fault in any active
component or system. Normally static components will not be considered to fail where adequate
protection from damage is demonstrated and reliability is to the satisfaction of the administration.
Single failure criteria include:
any active component or system (generators, thrusters, switchboards, remote controlled
valves, etc.)
any normally static component (cables, pipes manual valves, etc.) which is not properly
documented with respect to protection and reliability;
For equipment Class 3, a single failure includes:
items listed above for Class 2, and any normally static component is assumed to fail
all components in any one watertight compartment, from fire or flooding
all components in any one fire subdivision, from fire or flooding.
1.5.1 Key Elements of Fault Tolerant Systems
In DP vessel design, fault tolerance is generally achieved by redundancy. Redundancy is
defined as having more than one means of carrying out the same function. However the
effectiveness of redundancy is degraded if each method of carrying out the function is not
sufficiently reliable. The International Electrotechnical Vocabulary IEC 60050 defines
reliability and redundancy as follows:
Reliability The probability that an item can perform a required function under given
conditions for a given time interval.
Redundancy The existence of more than one means of performing a required function.
In fault tolerant systems based on redundancy there are three important elements that must
be present and maintained, these are:
equivalence;
independence;
confidence.
In practical terms these requirements can be translated to:
performance;
protection;
detection.
1.5.2 Equivalence
Redundant elements must be present in both number and capacity. That is to say that if
there are redundant means of performing a function then the secondary means should
provide the same level of performance as the primary means. If this is not the case then the
vessels post-failure capability will be determined by the performance of the secondary
system.
1.5.3 Independence
In practical fault tolerant systems based on redundancy there will be a number of common
points where redundant elements are linked together. There must be a comprehensive set
of protective functions designed to ensure that faults in one redundant element are
prevented from adversely affecting the performance of all other redundant elements
connected by that common point.

4 IMCA M 206
1.5.4 Confidence
Where fault tolerance depends on a backup system or standby redundancy, there will always
be a degree of uncertainty about the availability of the equipment when required. Alarms and
period testing are generally accepted as means of providing the necessary level of confidence.
Fault tolerance also depends on all systems being set up correctly. The correct configuration
of such things as crossover valves, backup electrical supplies and duty/standby pumps should
be recorded in the appropriate DP checklists.
1.6 DP Class 2 and DP Class 3 Redundancy Concepts
All DP Class 2 and DP Class 3 vessels must be able to maintain position and heading following any
defined single failure appropriate to their DP equipment class. The Worst Case Failure Design Intent
(WCFDI) defines the minimum amount of propulsion machinery which remains fully operational
following the worst case failure and therefore defines the vessels post-failure capability.
DP Class 2 and DP Class 3 vessels should work within their post-failure DP capability, when
conducting operations requiring a vessel of that equipment class so that position and heading can be
maintained following a failure.
The redundancy concept is the means by which the WCFDI is assured and describes how fault
tolerance is achieved on a system-by-system basis. The redundancy concept will be described in the
vessels DP system failure modes and effects analysis but may also be contained in a standalone
document or in the DP operations manual.
For DP Class 2 the redundancy concept will explain the effects of technical failures within each
subsystem of the overall DP system and describe any features and functions upon which fault
tolerance depends.
For DP Class 3 the redundancy concept will also describe how the DP system is able to maintain
position and heading following the loss of one fire subdivision or watertight compartment to the
effects of fire or flooding.
In the language of some classification societies these properties of fault tolerance are described as
redundancy in technical design for DP Class 2, with the addition of physical separation of redundant
elements for DP Class 3.
1.7 Failure Modes and Effects
Throughout the technical discussions that follow, reference is made from time to time to the ways in
which systems can fail. The term failure mode is used to indicate the manner in which a component
or subsystem fails. A faulty component may have several modes of failure, for example a cable may
fail, open circuit, short circuit or develop an earth fault. One component failure mode can have a
relatively benign effect while another may cause a blackout. Particular attention must be paid to
components that can fail in an active way and not just to an inert state. Generator control systems
are a typical example of systems that can fail in this way.
The term failure effect is used to describe the effect of that failure mode on the system itself and the
systems to which it is connected. The term end effect is sometimes used to describe the effect of a
particular failure mode on the vessels ability to maintain position and heading.
It is a classification society requirement that the fault tolerance of DP Class 2 and DP Class 3 vessels is
proven by a failure modes and effects analysis.
1.7.1 Typical List of Equipment and Functions Related to DP
Table 1 provides a typical list of the equipment and functions that have an impact on a DP
Class 2 or DP Class 3 redundancy concept. Not every vessel will have all the equipment or
functions listed but most modern diesel electric vessels will have something along these lines.


I
M
C
A

M

2
0
6


5


1 2 3 4 6 7 8 9 10

Engines and
Auxiliary
Systems
Power
Generation
Power
Management
Power
Distribution
Thrusters and
Main Propulsion
Vessel
Management DP Control Safety Systems
Separation for
Fire and Flood
1
Engines and
engine mounted
systems Alternators PMS architecture
Power
generation level
HV or LV Tunnel thrusters Topology
Main and backup
DP controllers
Fixed firefighting
systems Cable routes
2
Engine control
system and
shutdowns
Generator and
switchgear
control
Standby
selection
Service drilling
ROV diving
pipe-lay
Azimuth
thrusters
Operator
stations
Operator
stations Fire dampers Pipe routes
3
Fuel oil storage
transfer and
distribution
Generator
protection
Generator
start/stop
Major auxiliaries
380V 440V
690V Main propellers
Field stations
Remote I/O DP networks
Watertight
dampers Fire sub divisions
4
Lubrication
including pre-lub
Bus bar
protection
Active and
reactive power
sharing LV protection Gearboxes
Hubs and
switches
Vessel sensors
MRU, gyros,
wind, draught
gauges ESD
Watertight
compartments
5 Cooling SW Synchronising
Load dependent
start/stop
Service
transformers Rudders
Distribution
units
Position
references
DGPS, HPR,
TW, FB F&G
Watertight
doors
6 Cooling LTFW Governors Alarm start
Lighting, small
power
120V 220V Motors UPS for VMS
Manual thruster
controls Group E-stop
Compartment
analysis
7 Cooling HTFW AVRs
PMS blackout
prevention
Emergency
380V 440V Motor starters
Independent
joystick
8
Charge air
cooling/heating
Manual
switchboard
controls
Blackout
prevention in
other systems
Emergency
120V 220V
Variable speed
drives UPS for DP
9
Fuel valve
cooling
Blackout restart
and recovery
Switchboard
control supplies
Thrusters
auxiliaries
Voice
communication
10 Start air
Heavy consumer
control
Battery systems
24Vdc, 110Vdc
Thruster
shutdowns DP alert
11 Control air
Auto
reconfiguration
Wheelhouse
24Vdc
Thruster
emergency stops


6


I
M
C
A

M

2
0
6


1 2 3 4 6 7 8 9 10

Engines and
Auxiliary
Systems
Power
Generation
Power
Management
Power
Distribution
Thrusters and
Main Propulsion
Vessel
Management DP Control Safety Systems
Separation for
Fire and Flood
12 Service air
Control and
starting supplies
for emergency
generator
Speed torque
pitch control
13
Engine room
ventilation
UPS for other
systems Azimuth control
14
Remote control
valves Interlocks
15
Emergency
generator
Ventilation and
HVAC for
spaces other
than ER
Table 1 Typical DP related equipment and functions associated with the redundancy concept


IMCA M 206 7
2 Power Generation
2.1 General Principles of Power Generation
2.1.1 Voltage, Current and Power
When a voltage (V) is applied across an electric load it will draw a current (I) depending on
the impedance (Z) of the load, according to the relationship in Equation 1. For alternating
current circuits, V, I and Z are complex quantities.
IZ V
Equation 1 Ohms law
Impedance is measured in Ohms () and may be resistive, reactive or a combination of these.
Reactance (X) itself may be inductive (X
L
) or capacitive (X
C
) such that impedance is the
algebraic sum of resistance and reactance.
Resistance (R) is also measured in Ohms;
Inductance (L) is measured in Henrys (H);
Capacitance (C) is measured in Farads (F).
By convention, inductive reactance is considered to be positive and capacitive reactance is
considered to be negative. Figure 1 shows the impedance triangle which is one of the core
principles of electrical engineering.
jX
L
RESISTANCE AND
INDUCTIVE REACTANCE

R
Z
-jX
C
RESISTANCE AND
CAPACITIVE REACTANCE

R
Z
Z = IMPEDANCE IN OHMS ()

Figure 1 Impedance as the combination of resistance and reactance
Where the applied voltage is sinusoidal, the current drawn by the load is displaced with
respect to the applied voltage depending on the combination of resistance and reactance.
A load consisting of purely capacitive reactance draws a sinusoidal current which leads the
applied voltage by 90. Similarly, a load consisting of purely inductive reactance draws a
current which lags the applied voltage by 90. Where the load has an impedance containing
both resistance and reactance the current will lead or lag the applied voltage depending on
the relative amounts of each. Figure 2 shows the current and voltage relationship for a
resistive load. Figure 3 shows the same for a load with some resistance and some inductive
reactance such that the current waveform lags the applied voltage waveform by 30.
The magnitude and angle of an impedance can be calculated using Equation 2:

8 IMCA M 206
R
X
Z X R Z arctan
2 2

Equation 2 Impedance
0 5 10 15 20
-10
-5
0
5
10
Time (ms)
V
o
l
t
a
g
e
-
R
E
D
,

C
u
r
r
e
n
t
-
B
L
U
E


Figure 2 Current and voltage in phase (purely resistive load)
0 5 10 15 20
-10
-5
0
5
10
Time (ms)
V
o
l
t
a
g
e
-
R
E
D
,

C
u
r
r
e
n
t
-
B
L
U
E


Figure 3 Current lagging voltage by 30 (slightly inductive load)
The instantaneous power at any point in time is the product of the instantaneous voltage and
current. Figure 4 shows that power flow is always positive from the resistive load. This
represents fuel from the generators fuel tank being turned into mechanical power and waste
heat. On the other hand, power flow goes negative in Figure 5, indicating that a certain
amount of power is being exchanged between the generator and the load. Note that in the
case of a thruster motor powered by a variable speed drive with a rectifier frontend, the
power interchange is between the motor and the drive and thus the generator sees the
power factor of the drive and not the motor.

IMCA M 206 9
0 5 10 15 20
-2
0
2
4
6
8
10
12
Time (ms)
I
n
s
t
a
n
t
a
n
e
o
u
s

P
o
w
e
r

Figure 4 Instantaneous power for a purely resistive load
0 5 10 15 20
-2
0
2
4
6
8
10
Time (ms)
I
n
s
t
a
n
t
a
n
e
o
u
s

P
o
w
e
r

Figure 5 Instantaneous power for a slightly inductive load
The displacement of voltage and current gives rise to the concept of power factor.
The power factor of a load, cos , can be defined in several ways such as the ratio of the
apparent power (product of V and I) to the active power (kW), or as the cosine of the angle
between the voltage and current. In the case of Figure 3, the power factor is cos 30 =
0.866.
Almost all loads draw a combination of active (P) and reactive power (Q). The active power
is measured in Watts or Kilowatts and can be thought of as useful work and waste heat. This
requires the consumption of fuel. The reactive power is measured in Volt Amps reactive or
kVAr. It is used to maintain the electric and magnetic fields in the cables and electric
machines and does not require fuel consumption. However, the reactive currents do give
rise to heating and additional losses. Together, the active and reactive power combine
algebraically to give the apparent power and the power factor of any load, as show in Figure
6.

10 IMCA M 206
(kVAr) REACTIVE POWER
STORED IN ELECTRIC
AND MAGNETIC FIELDS

APPARENT POWER
(kVA)
REPRESENTS REAL POWER DELIVERED AS
MECHANICAL POWER OR HEAT.
REPRESENTS CONSUMPTION OF FUEL
ACTIVE POWER (kW)
POWER FACTOR
(COS )

Figure 6 Active, reactive and apparent power
Power factor can vary between 0 and 1, with 0 representing an entirely reactive load and 1
representing an entirely active (resistive) load. Non-adjustable loads such as heaters will have
unity power factor. Pumps and fans will usually have a power factor in the region of 0.85 at
full load but it may be considerably lower at partial load. Loads which can vary the amount of
active power they can draw from the generators (such as thruster motors) can have a
variable power factor which improves as they draw more active power. Although reactive
power does not directly equate to fuel consumption, the fact that additional generators need
to be online to supply the reactive power means that there are additional losses associated
with the inefficiency of running another generator just for this purpose. Power factor
correction is the term applied to schemes intended to supply reactive power from sources
other than generators to improve overall plant efficiency.
Particular care is required in applying power factor correction methods to marine power
systems. In shore-based applications it is common practice to correct poor power factor by
connecting static capacitor banks. In marine power systems this can have unexpected effects
including creating system resonance leading to severe over voltage and equipment damage.
The reliability of capacitors can also be an issue if they short circuit, creating a severe voltage
dip on the power system. Other methods of power factor correction include synchronous
condensers and active VAr compensation.
2.1.2 Generating Electricity
There are several ways of generating electricity, such as fuel cells, piezo-electric effects and
chemical reactions, but the method employed for large scale power generation involves
moving a conductor (or system of conductors) through a magnetic field or moving a magnetic
field past a system of conductors, as shown in Figure 7. This principle is employed in AC and
DC power conversion and is governed by Equation 3 and Equation 4 where:
E is the electro motive force generated in the conductor in Volts;
B is the magnetic flux density in Tesla;
I is the current in amperes;
l is the length of the conductor in metres;
v is the velocity of the conductor in metres per second;
F is the force in newtons.
Note that a force will be created on a stationary current carrying wire sitting in a magnetic
field (motor effect), as shown in Figure 8. However, to induce an electro motive force
(voltage) in a wire and therefore drive a current through a load, the wire must be moving
(generator effect).

IMCA M 206 11
Blv E
Equation 3 EMF
BlI F
Equation 4 Force
MAGNETIC FIELD B
S
GENERATOR EFFECT
MOTION
I
l
E
N

Figure 7 Conductor passing through a magnetic field induces an EMF

FORCE
MAGNETIC FIELD B
S
I
l
N
MOTOR EFFECT

Figure 8 Current carrying conductor in a magnetic field experiences a force
2.1.3 Alternators
Almost all diesel electric DP vessels use variations on the three-phase, brushless, self-exciting,
synchronous alternator as the means of generating electricity. The alternator converts
mechanical energy from the diesel engine into electrical energy at nominally constant voltage
and frequency. The frequency and voltage produced by the alternator are determined by
design parameters, but for a given design the frequency of the output waveform is controlled
by varying the speed of the diesel engine and the voltage by controlling the current in the
rotor winding. The power delivered by the alternator is controlled by varying the fuel
admission to the engine. There are two dedicated control systems for this purpose, the
engine governor and the automatic voltage regulator. When generators of this type operate
in parallel with each other, all connected machines naturally run at exactly the same speed (or
related speed if different number of poles) and are said to be synchronised together.
Figure 9 shows the cross-section of a salient pole alternator.

12 IMCA M 206
POLE PIECE
STATOR WINDING (AC)
STATOR
ENCLOSURE
SALIENT POLE ROTOR
ROTOR WINDING
S
N
S
N

Figure 9 Cross-section of four pole, salient pole alternator
The three-phase winding is located on the stator (or stationary part) and typically configured
as a wye (or stars), as shown in Figure 10. The rotor winding is located on the rotating part
and arranged to create the required number of magnetic poles. On a salient pole alternator
(as opposed to one with a cylindrical rotor) the rotor poles are easily identified and each is
capped with a dedicated pole piece. On brushless alternators the rotor winding is connected
to a small AC generator (exciter) mounted on the same shaft by way of diodes. The
stationary part of the exciter winding is powered from the automatic voltage regulator.
The stator winding may have a three or four wire connection depending on whether the
neutral point (or star point) of the wye is used. The terminals of the machine are often
annotated red, yellow and blue or U V W. Note that it is possible to specify an alternator
with a clockwise or anticlockwise phase rotation and this needs careful consideration in
relation to an engines direction of rotation. Failure to properly specify the direction of
rotation can result in problems in the distribution system including motors running the wrong
way.

STATOR
RED
YELLOW
NEUTRAL
BLUE
ROTOR

Figure 10 Three-phase alternator with wye (star) winding
The stator winding is arranged to create a 120 phase shift between the three voltage
waveforms produced. Figure 11 shows the phase voltage waveforms produced by an 11kV,
60Hz alternator.

IMCA M 206 13
0 5 10 15 20
-10
-5
0
5
10
Time (ms)
V
o
l
t
a
g
e

(
k
V
)

Figure 11 Voltage waveforms for red, yellow and blue phases of 11kV alternator
The alternators nameplate voltage rating is called the line voltage and refers to the root
mean square (RMS) voltage between the terminals; for example red to blue, blue to yellow
and yellow to red. The term phase voltage refers to the voltage developed across each
winding, such as red to neutral, yellow to neutral and blue to neutral. The phase voltage and
the line voltage are related as shown in Equation 5. The line currents are equal to the phase
currents in a wye winding.
Vphase Vline 3
Equation 5 Relationship between line and phase voltage in wye winding
Equation 6 can be used to determine the engine speed required to produce the desired
frequency for an alternator with p poles. Table 2 lists the engine speeds required to
produce 50Hz and 60Hz from four, eight and ten pole alternators.
p
f
n
120

poles of number p and frequency f minute per s revolution in speed n where ,
Equation 6 Relationship between speed, frequency and number of poles
Engine Speed (rpm) Frequency (Hz) Number of Poles
1800 60 4
1500 50 4
900 60 8
750 50 8
720 60 10
Table 2 Typical engine speeds and pole arrangements for 50Hz & 60Hz alternators
The basic, per phase, equivalent circuit of an elementary three phase synchronous alternator
operating under steady state conditions is shown in Figure 12. E represents the electro
motive force (EMF) created by the excitation system and Xs is the internal impedance, called
the synchronous reactance, represented by an inductance. V is the voltage across one
winding (phase voltage) at the terminals of the machine. In practice, the windings also have
resistance but this is generally neglected for ease of comprehension.

14 IMCA M 206
jX
S
V E
I
L
I
L
X
S

Figure 12 Elementary equivalent circuit of synchronous alternator
The phasor diagram in Figure 13 shows the currents and voltages when the generator is
delivering power to a lagging (inductive load) as is typical for a marine application.
L
I
jI
L
X
S

V
E


Figure 13 Phasor diagram
The phasor sum of the terminal voltage (V) and the voltage drop across the internal
impedance (I
L
X
S
) add to give the EMF of the machine. The angle of the internal voltage drop
depends on the power factor (cos ) of the load and together with the magnitude of the
current being supplied, determines the load angle () of the generator. In theory, the
alternator will stay synchronised for load angles up to 90 but in practice, the stability limit of
the machine will be much lower. Large voltage and current swings can occur if a machine
loses synchronism with other machines when the practical limit is exceeded.
2.1.4 Generator Capability
Several other design features determine the operating limits of the generating set and it is
common practice to develop a generator capability plot which shows the steady state limits
of machine operation as shown in Figure 14. The plot takes the form of a P and Q axis with
the first quadrant representing power flow out of the generator and lagging reactive power.
In most practical applications, the upper limit is determined by the power that can be
delivered by the engine. The limits of positive reactive power are defined by the thermal
rating of the windings. The limits of negative reactive power form the practical limit of
stability. Parallel operation beyond this limit risks the machine breaking synchronism leading
to severe voltage and current transients with the potential for blackout. In practical
operations the machine will almost always operate in the first quadrant. How the alternator
is protected against the effects operating out with this envelope is discussed in section 2.12.

IMCA M 206 15
(kW)
kW LIMIT
IMPOSED BY ENGINE

1
ST

QUADRANT
LOAD ANGLE
THEORETICAL
STABILITY LIMIT
WINDING THERMAL
LIMIT
NAMEPLATE
POWER FACTOR
REACTIVE POWER
(kVAr)

Figure 14 Generator capability plot
2.1.5 Alternator Nameplate Power Factor and Operating Power Factor
The nameplate power factor of an alternator is just one of the parameters used to define the
power capability of a generating set. Electrical standards typically state that the output rating
of the generator will be given in kVA at rated voltage, current, frequency and power factor.
Alternators are rated in kVA or MVA but the amount of active power (kW) they can deliver
is determined by the power of the engine driving the alternator. The amount of current they
can deliver is determined by winding construction and cooling etc.
The nameplate power factor (cos ) is the ratio of the real power (P in kW) delivered to
the apparent power (S in kVA) when the generator is operating at rated current, voltage and
frequency, cos = P/S.
Typical nameplate data:
5.375MVA = S = apparent power;
11kV = V
LINE
;
282A = I
LINE
;
0.8 p.f. (Cos ).
From the information above it is possible to calculate the rated output power by using
Equation 7.
3 Cos I V P
LINE LINE

Equation 7 Three phase power
The mechanical power into the shaft of the alternator must be higher than the electrical
power output so the example alternator would typically be driven by an engine capable of
slightly more power than this to allow for losses such as friction, windage and waste heat.
The nameplate power factor indicates how much reactive power (kVAr) the generator can
supply continuously without overheating when it is delivering rated active power (kW).
In the case of the example generator 4.3MW/5.3MVA = 0.8. The nameplate power factor
operating point is just one point in the generators operating range.
The operating power factor is the ratio of the real power (kW) to the apparent power (kVA)
that the generator is actually supplying to any load and changes as the load conditions change,
so it can be any value from 0 to 1, leading or lagging (loads are mostly lagging).

16 IMCA M 206
The generator can supply loads at any operating power factor within its capability curve.
If the operating point goes outside the capability curve the alternator may trip on over
current or some other protective function. It may also become unstable and lose
synchronism with the network if the operating point swings across the stability limit.
However, regardless of nameplate details the generator must accept whatever power factor
is demanded by the load, as shown in Figure 15. In the case of parallel operation of two or
more generators, each generator can be arranged to provide different portions of the kW
and kVAr demand (normally set to be equal).
(W) ACTIVE POWER
W LIMIT
IMPOSED BY ENGINE
1
ST

QUADRANT
LOAD ANGLE
PRACTICAL
STABILITY LIMIT

WINDING THERMAL
LIMIT
OPERATING
POWER FACTOR
REACTIVE POWER
(VAr)
OPERATING
POINT
3MW
1.5 MW
3MVAr
1.0MVAr

Figure 15 Generator operating point
2.2 Engines
2.2.1 Marine Engine Types
Engines for electric power plant are almost invariably four stroke, turbocharged medium
speed diesel engines. Some applications use two stroke diesels, and others such as those for
FPSOs make use of dual fuel (MDO or fuel gas) and some applications use gas turbines.
Marine diesel engines are categorised by their speed. Table 3 shows the speed range and
traditional application in DP vessel power plant. Slow speed engines are used primarily in
merchant vessel applications and usually only found in DP vessels that have been converted
from an existing merchant ship such as a tanker.
Category Speed range Application
High speed >1000 rpm Auxiliary and emergency generators
Medium speed 400-1000 rpm Main, auxiliary and emergency generators
Main propulsion engines
Slow speed < 400 rpm Main propulsion engines
Table 3 Diesel engine speed and typical use
Typical medium speed diesel engines for DP vessel applications are rated from around 1MW
in small vessels to 10MW in large vessels. Installations of four, six or eight engines are
commonplace with 2MW to 7MW being a popular power range. The engines are invariably
multi-cylinder units in either in-line or V configuration, as shown in Figure 16 and Figure 17.

IMCA M 206 17

Figure 16 MAN 8L 32/40 Courtesy MAN 32/40 Project Guide

Figure 17 MAN 16V 32/40 Courtesy MAN 32/V40 Project Guide
2.2.2 Engine Rating
Diesel engines for generators are rated in kW at synchronous speed. The engine
manufacturers data sheets may provide information such as:
Cylinder bore 320mm;
Piston stroke 400mm;
Cylinder output 500kW/cyl;
Speed 750rpm;
Mean effective pressure 24.9bar.
For an HV generator, the electrical power output is typically around 3% less than the power
rating of the engine, to allow for the losses in the electrical machine. In the case of vessels
with different sizes of engines such as in a father-son arrangement, the same alternator may
be fitted to all the engines such that the smaller engine is rated well below the alternator
rating. Care should be taken when specifying engine and generators of different sizes.
Transient stability may require closer scrutiny as the difference in generator ratings
approaches 2:1.
2.2.3 International Maritime Organization (IMO) Nitrogen oxide (NOx) Emissions
The performance of modern diesel engines is influenced by the need to comply with IMO
requirements for exhaust gas emissions described in Annex VI of the International
Convention of the Prevention of Pollution from Ships (MARPOL) MARPOL 73/78. Engines

18 IMCA M 206
complying with these requirements are described as IMO Tier I. Further reductions in NOx
emissions to be designated IMO Tier II and Tier III become effective for ships built on, or
after, 1 January 2011 and 2016 respectively.
2.2.4 Speed Regulation
Diesel engines are fitted with a speed governor, which will control the engine fuel admission,
so as to maintain the desired speed (and load in the case of parallel operation of generators)
within the engines ability to accept the load change. In diesel electric power plants it is
possible for very large load changes to occur in normal operation, or in response to faults in
the power system, and the plant designers should ensure that measures are in place to limit
the magnitude and rate of change of load to levels within the capability of the engines. The
ease with which the power consumption of large variable speed drives can be controlled has
greatly improved the management of such load changes.
The major classification societies are fairly well aligned on requirements for the speed
regulation of engines for main and emergency power. In general:
Transient frequency variations will not exceed 10%;
Steady state frequency must be achieved within 5s of the maximum permissible step load
being applied or thrown off;
Transient frequency variations in excess of 10% are accepted in the case of 100% load
rejection provided the generator does not trip on over speed;
For diesel generators, the over speed protection will shut down the engine at 115%.
2.2.5 Engine Start-up Time
Generators using high speed and smaller medium speed engines are generally able to start
and achieve nominal speed within eight to ten seconds. Larger medium speed engines can
reach nominal speed in around 20 seconds, but some larger medium speed engines may take
longer and have special requirements in respect of pre-lubrication, length of time on standby
and the need to perform a slow turn before starting. The run up time may also be controlled
to limit the acceleration and therefore the amount of smoke generated during starting.
Figure 18 shows the starting time for a large medium speed diesel engine starting from the
not on standby condition. If the engine needs to be pre-lubricated it could take over two
minutes to connect the generator. If pre-lubrication was running at the time of the start
signal then the connection time would begin with the slow turn function and reduce to
around one and a half minutes. If the generator was in hot standby mode then the slow turn
would be omitted and connection times reduced to around one minute. The time taken to
synchronise the generator may depend to some extent on the stability of the power system
at the time of connection.
It may be possible to improve upon standard starting and connection times with the engine
manufacturers agreement and assistance. Designers should enquire about starting
restrictions as these can significantly influence time taken to connect standby generators in
response to engine failure. This may in turn influence the power plant operating strategy in
respect of the amount of spinning reserve that must be maintained.
Blackout recovery times can also be adversely affected by starting requirements which should
be minimised as far as possible for DP vessels.

IMCA M 206 19
0 20 40 60 80 100 120 140 160
0
20
40
60
80
100
Time (s)
S
p
e
e
d

(
%
)

Prelubrication
Slow
Turn
Run Up

Sync

Start Signal

Figure 18 Medium speed engine starting time
2.2.6 Restriction on Standby Status and Starting Time
Some diesel engines have restrictions on the time they can remain on standby without an
increase in the starting time. These restrictions can be due to such things as oil build-up due
to pre-lubrication which would need to be removed. Power management systems may
include engine management routines specifically to deal with these features. Typical
restrictions may include:
Engine to be at preheat temperature and pre-lubrication to be running;
Up to twelve hours engine may start without slow turn. This condition is sometimes
referred to as hot standby;
After twelve hours of pre-lubrication, engine needs to perform a slow turn before
starting, which extends starting time. This condition is sometimes referred to as cold
standby.
Management of these restrictions typically includes the power management system
monitoring the time an engine has been on standby. As the time approaches twelve hours,
the PMS will start the engine and run at 30% power or above. If this is not possible the
engine status will be changed to not standby.
2.2.7 Black Start Capability
Most engines are capable of being started in an emergency situation without pre-lubrication,
providing the engine jacket water is at or above a defined temperature and the pre-
lubrication was running a short time before the start request. A time limit may be set for the
maximum time after which the engine can be started without pre-lubrication. Such limits can
be of the order of one minute, which should be adequate for most DP blackout recovery
purposes. Confirmation should be sought from the engine manufacturer that the engine can
be started in this way. It may be necessary to provide a dedicated black start signal to the
engine control system. Although it is good practice to design the blackout recovery system
to be independent of the emergency generator, it is prudent to ensure that all systems
required to black start the engines can be supplied from the emergency generator in case the
engines cannot be started within the pre-lubrication override time. A dual supply to such
consumers from the main and emergency power system is recommended, with the main
power system providing the normal source of power.

20 IMCA M 206
Where an exemption from pre-lubrication is not possible or as additional protection, an air
driven pre-lubrication pump can be driven from the starting air system.
Where the engine requires fuel pressure for starting it may be possible to arrange a gravity
feed tank or an air driven fuel pump to be activated for blackout recovery.
2.2.8 Load Application Time
Most engine manufacturers will provide figures for the maximum continuous rate of
application of load (kW/s) and for the maximum step load that can be applied. In DP
applications, the rate of load application required by the thrusters for accurate station
keeping may approach the limits of the engines capability and plant designers should ensure
that loading ramps are not unnecessarily severe. Most thruster control systems can be
programmed to accept a loading ramp to ensure limits are not exceeded. Similar safeguards
should be applied to large non DP related consumers such as active heave compensation or
drilling consumers.
Load increase on an engine may occur due to increasing demand from consumers, but also as
a result of engine failure leading to loss of generating capacity. In many DP power plant
designs it is possible for several generators to be lost together as the result of a single failure,
therefore plant designers need to consider the maximum loss of power generating capacity
that is likely to occur, not just failure of one generator. Rapid load shedding functions may
assist in meeting engine manufacturers requirements.
0 20 40 60 80 100 120 140 160 180
0
10
20
30
40
50
60
70
80
90
100
Time(s)
P
o
w
e
r
(
%
)

Blackout Normal
Standby
DP

Figure 19 Load increase rates for generators operating at synchronous speed
Engine manufacturers may also define an emergency or blackout load up ramp which is
considerably steeper than the normal ramp.
In Figure 19, the load ramp required by thrusters, i.e. DP, falls between the normal operating
curve and blackout curve.
Note that load application may be the combination of load increase on the overall system and
the load transfer to a newly connected generator which is following a load up ramp.
2.2.9 Load Steps
At one time classification societies required generators to be capable of a 50% load step
without unacceptable fall in frequency. Many modern medium speed diesels are highly turbo
charged and cannot achieve these levels of step load because of the lag introduced by the
turbocharger. Current practice is to make reference to IACS requirements or ISO 8528-5
which can be achieved. Engines which meet these requirements may be referred to as 3 or 4
step engines, that is to say they require that the largest instantaneous load step be limited to
some specified value. Some manufacturers simply state a maximum step load such as 33% of
MCR, whereas others relate the maximum permissible step load to the load on the engine at


IMCA M 206 21
the time the step is applied, as shown in Table 4. Limits may include a period of stabilisation
after application of the step load which may be of the order of ten seconds.
% MCR 0 33 56 74
Max load step % 33 23 18 26
Table 4 Maximum permissible step load
To reduce the effects of turbocharger lag and thus improve the step response some
manufacturers employ other methods such as a combination of mechanically/electrically
driven compressors and exhaust gas driven compressors. Driving the turbocharger
compressor with a short blast of compressed air is another method used to improve step
response.
2.2.10 Load Rejection
The engine manufacturer may specify the maximum load rejection rate for normal operation
but in diesel electric designs the possibility of the generator circuit breaker opening with the
engine at or above full load should be considered.
In cases where the generator remains connected to the power distribution system and the
load rejection is caused by tripping of consumers, medium speed diesels can generally accept
instantaneous 100% load rejection without operation of the over speed device, but speed
deviation may be higher than that specified for normal operation.
For power system operating in uncorrected droop the requirement to accelerate a large
number of pumps and fans from low frequency to high frequency can act as a buffer to limit
the rate of load rejection.
2.2.11 Low Load Running
The following terms are used by engine manufactures to describe engine loading:
Overload > 100% rated power;
Full load = 100% rated power;
Part load < 100% rated power;
Low load < 15% (MDO) 25% (HFO) rated power;
No load = 0% rated power.
Long periods of low load running can be a feature of DP operations particularly where the
fault tolerance of the power plant depends on maintaining a large spinning reserve. As this
type of operation is detrimental to the engine, efficiency, emissions, performance and
reliability, some engine manufacturers state a minimum figure such as 15% of MCR if the
engine is operated on diesel fuel oil.
If the DP power plant has been designed to have a low impact worst case failure (WCFDI = 1
generator) then the plant can generally be run at relatively high loads. However, if prolonged
low load running is unavoidable there are a number of features that can be considered to
improve conditions:
Most power management systems offer load dependent stop as a way to manage low
load running but most operators prefer to use this in an advisory mode;
Most power management systems offer an asymmetric load sharing function to allow
soot deposits to be burnt off on one engine at a time;
Engines can be fitted with part load nozzles;
Engines can be provided with two stage charge air cooling systems;
Some high speed diesels can reduce the number of cylinders in use at low power output
levels.

22 IMCA M 206
2.3 Engine Auxiliary Systems
2.3.1 General
The term marine auxiliary systems is used to describe all those systems and functions
required to support thruster, generator and related plant. Some of the systems may be
mounted on the generator skid or as standalone units while others may be part of the engine
room and pump room design. In any fault tolerant system based on redundancy it is the
common points between otherwise separate and redundant elements of the design that allow
faults to propagate from one redundant element to the other. This is particularly true of
engine auxiliary systems related to fuel, compressed air and cooling water systems where
several generators may be connected to a common system. In recent years some
classification societies have introduced additional DP rules related to the design of marine
auxiliary systems to try to limit the potential for more than one generator being lost as the
result of a single failure. There are significant advantages to making generators as
independent as possible even if power distribution system faults could cause the loss of
multiple generators.
2.3.2 Fuel
Most engines for DP vessels use diesel fuel oil but heavy fuel oil systems can be found in some
applications such as heavy lift vessels and shuttle tankers. Heavy fuel oil systems need careful
attention to ensure there are no single point failures associated with fuel heating systems or
changeovers from HFO to DFO.
A secure supply of clean fuel is essential for reliable operation. Although redundancy in DP
system design generally begins at the fuel oil day tanks, reliability begins with the design and
operation of the fuel oil storage, purification and distribution systems. Fuel system faults are
generally associated with blockages, leaks and contamination by water or MBC.
Although DP Class 2 failure definitions allow redundant generators to share common pipe
work, some classification societies now require that fuel systems for engines, considered to
provide redundancy, are totally separate. Crossover valves are permitted for maintenance
but they must be kept closed during DP operations, requiring the vessel to be fault tolerant.
The fuel system design benefits from fuel oil pressure alarms, differential pressure alarms at
filters, water detectors at purifiers and leak detectors for high pressure fuel lines. Provision
of independent tank level gauges and low level alarms offer additional protection against an
empty day tank.
The practice of running ballast or cooling water lines through fuel storage tanks should be
avoided. Faults in engine fuel oil coolers can also be responsible for water contamination.
Where engine operation depends on fuel pumps, consideration should be given to how a fuel
supply can be provided for blackout recovery. A limited supply from a gravity tank or
pneumatically operated black start fuel pump are amongst the options.
In general the design of the engine fuel supply system should follow the overall split in the
redundancy concept including day tanks, supply and return lines, filters, pumps, quick closing
valves and their controls.
2.3.3 Lubricating Oil
Engines generally have independent lubrication systems but engine room pipe work intended
to provide facilities for replenishing engine sumps and recovering waste oil may increase the
risk of inadvertently emptying one engine and simultaneously overfilling another.
Consideration should be given to the provision of valves and control measures to prevent
such acts of mal-operation.

IMCA M 206 23
2.3.4 Starting Systems
Starting system for diesel engines can be by electric or pneumatic starter motor or by
distributor type starting where high pressure air at 30bar is injected directly into the engine
cylinders. The latter is the most common system on large medium speed engines.
The starting air system should be designed to meet classification requirements for the type of
vessel in relation to numbers of starts on each engine etc. Where fault tolerance depends on
the successful connection of standby generators the starting system needs to be particularly
robust.
The design of the starting air system should also consider the blackout restart strategy.
In some designs the power management system will simultaneously order the start of all
available generators following detection of blackout. In such cases, the starting air system
should be designed to allow simultaneous cranking.
Where low pressure control air supplies are derived from high pressure starting air systems,
a risk of overpressure exists if the pressure reducer fails in such a way that it passes high
pressure air to the lower pressure system. Overpressure in the control air system can cause
unexpected effects including the operation of engine stop cylinders. Correctly calibrated
pressure relief valves provide a degree of protection as does the relative volume of the lower
pressure system if it is large in comparison to the high pressure systems.
2.3.5 Control Air
Control air systems for engines can have a wide variety of uses including:
oil mist detectors;
start and stop cylinders for the fuel system;
operation of rig saver combustion air shut off valves;
operation of combustion air dump valves;
temperature control valves for HTFW and LTFW systems;
jet assist for the turbocharger;
pneumatically operated governors.
Problems with the design of control air systems generally arise when failure of the control air
supply has an adverse effect on engine operation and several engines share a common control
air supply.
At least one classification society now requires full separation of pneumatic systems serving
equipment intended to provide redundancy even for DP Class 2 designs.
2.3.6 Combustion Air and Engine Room Ventilation
For DP Class 2 and DP Class 3 vessels ventilation and the supply of combustion air to the
engines is considered in the same way as any other marine system. No single failure should
cause the engines to be adversely affected to the point where position and heading cannot be
maintained. Although DP Class 2 permits designs having a single engine room, most modern
DP Class 2 vessels have at least two engine rooms making it easier to split ventilation and fire
control measures along the lines of the redundancy concept.

24 IMCA M 206
RIG
SAVER
VENTILATION
SUPPLY
DUMP VALVE
TURBO
GEN
EXHAUST
FAN
AIR OUT
COMBUSTION
SUPPLY
AIR IN
ENGINE
FIRE DAMPERS
LOUVRES
TO
FUNNEL

Figure 20 Engine room combustion air supply and ventilation
Combustion air can be drawn from inside the engine room or from outside the vessel by way
of dedicated combustion air vents connected to the turbocharger inlets. Figure 20 shows the
conventional arrangement where combustion air is drawn internally. A flap can be arranged
to draw combustion air from inside the engine room to ease starting in cold climates.
Ventilation is primarily concerned with the extraction of heat and fumes to maintain
acceptable engine room conditions. Engine manufacturers may stipulate a range of
permissible combustion air temperatures such as 15C to 35C with an occasional increase
to 45C. Vessels for Arctic service may require heating of combustion air in some
circumstances. Sources of air for ventilation and combustion should be arranged to avoid sea
spray, dust and exhaust fumes in all wind directions and vessel headings. Control of engine
room internal pressure is important for safety reasons and a slight overpressure is often
recommended.
Engine room ventilation and combustion air supply may be arranged separately with dedicated
combustion air fans directing air to the turbocharger inlets and ventilation being ducted to
locations for heat extraction.
Ventilation and combustion air failures can be related to:
failure of ventilation fans or their supplies;
closure of fire or watertight dampers;
operation of combustion air shut-offs on engines;
operation of combustion air dump valves on engines;
release of firefighting medium.
Experience suggests that fan failure does not usually affect engine operation in the short-term
but temperatures may rise relatively quickly. See also crank case breathers in Section 2.3.9.
Design problems may arise in respect of fault tolerance when fire damper operation is linked
to fan running status such that the fire dampers close when the fans stop. Closure of fire
dampers on a running engine room has a more severe effect and may result in significant
negative pressure in the engine room. This can represent a safety hazard and this should be
taken into consideration when conducting DP FMEA proving trials. Engine room fire
dampers should generally fail open or as set on loss of actuator power or control signal,
but classification society requirements may also influence such designs.

IMCA M 206 25
Poor design of F&G and ESD systems has also been known to cause fire dampers in
otherwise independent engine rooms to close because of power and communications faults in
safety system field stations. Poor choice of fail safe settings can also have this effect.
Closure of combustion air shut-offs (rig savers) results in immediate stop of the engine and
tripping of the generator on reverse power. Rig savers are normally designed to require air
pressure to close, and failures are generally related to spurious activation. Design problems
in relation to fault tolerance can arise where engines share a common control system for the
rig savers.
Combustion air dump valves (or flaps) are located after the turbocharger and are designed to
vent the air from the turbocharger into the engine room when the rig saver closes. Failure of
this valve to the open position results in a large reduction in the engines ability to generate
power. This creates a severe load sharing imbalance accompanied by large amounts of black
smoke from the exhaust because the governor advances to the full fuel position in order to
try and restore power output. This may distract attention from the true cause of the failure.
Design problems can arise in respect of redundancy when these valves fail to the open
position on loss of air supply and multiple engines share a common control air supply or
control system.
Classification societies may require that rig saver valves are located after the turbocharger in
view of the risk of over speeding of the turbocharger if the combustion air supply is cut off.
It has been stated that using combustion air from outside the vessel provides protection
against release of unexpected firefighting medium. This is an unlikely failure mode for
established firefighting systems but has occurred.
2.3.7 Jacket Cooling Water and Control
Most engines for DP vessels are fresh water cooled by way of heat exchangers served by the
vessels seawater cooling system. On smaller DP vessels engines may be radiator cooled or
by way of box coolers set into the hull. In some cases equipment may be directly cooled by
the seawater system. Circulation of coolant may be provided by engine driven pumps or
electrically driven pumps or by a combination of the two, as shown in Figure 21.
EXPANSION
TANK
ALTERNATOR
TT
ENGINE
JACKET
M
LUBE OIL
COOLER
CHARGE AIR
COOLER STAGE 2
CHARGE AIR COOLER
STAGE 1
PREHEATER
ENGINE
DRIVEN PUMP
ELECTRIC
PUMP
TT
SW
HEAT
EXCHANGER
ELECTRIC
PUMP
ENGINE
DRIVEN PUMP
36C
90C
LT
HT
LT

Figure 21 Engine HT and LT cooling water system

26 IMCA M 206
Engine cooling system may be divided into:
high temperature fresh water cooling (HTFW) typically 90C;
low temperature fresh water cooling (LTFW) typically 36 C.
These systems may in fact share the same cooling water and expansion tank by way of three-
way mixing valves designed to regulate the relative system temperatures.
The HTFW circuit will normally be connected to such things as:
first stage charge air cooler;
engine water jacket;
jacket water pre-heater;
water maker;
water maker booster heater.
The LTFW circuit will normally cool such things as:
second stage charge air cooler;
lub oil cooler;
alternator cooler.
Control valves for regulating engine temperature are generally:
wax element type AMOT;
electric actuator with electric controller motorised valve;
pneumatic actuator with electric controller.
Control valves are generally designed to fail to full cooling. Fail as set may not be sufficient
to prevent temperature instability in the power plant. Wax element type valves may fail to
open on rising temperature but the use of multiple elements means that failure is usually
gradual, providing an opportunity for the degradation of performance to be noticed.
Where several engines share a common LTFW system controlled by a single valve, failure of
the valve to open on rising temperature due to mechanical faults may be enough to cause
generators to trip even if the valve normally fails to the open position.
Temperature control may be by way of dedicated PID controllers, the engine manufacturers
control system or by the vessel management system.
Problems with the design of cooling systems in respect of fault tolerance usually arise when
designers interconnect generator cooling water systems to a common engine room LTFW
system or to supply a single water maker.
Design errors also include failure to correctly assign power supplies for pumps, temperature
control valves and their controllers in line with the overall split in the redundancy concept.
At least one classification society now requires that that FW cooling systems for engines and
other equipment intended to provide redundancy are totally separate (even for DP Class 2)
in view of the risk of leaks leading rapidly to overheating.
2.3.8 Jet Assist
Jet assist is a function added to highly turbocharged medium speed diesels to reduce the
turbo lag and increase the step response. Nozzles within the compressor casing direct air
onto the impeller to accelerate it rapidly in response to load changes. Jet assist may be
activated in response to:
engine starting;
generator circuit breaker closing;
speed undershoot;
sudden increase in fuel racks.

IMCA M 206 27
Jet assist time can be varied according to demand five to ten seconds is typical.
The activation and duration of the jet assist function can be controlled by the engine
manufacturers control system or by the vessel management system and requires information
on the speed of the engine, the fuel admission and starting air pressure. A number of
conditions are monitored and may cause the jet assist to be inhibited to prevent it being on
for too long or depleting the air supply.
Experience suggests that failure of the jet assist function to operate in response to step loads
does not cause generators to trip but may cause frequency based load shedding to function
briefly in extreme cases. Nevertheless, air supplies and control systems for jet assist
functions should be split along the lines of the redundancy concept.
Information from one engine manufacturer of large medium speed diesels suggests that
pulsating load changes of greater than 25% can occur as frequently as 30 times per hour in
DP vessels.
Design errors in relation to jet assist functions are normally related to under-sizing the air
supply and introducing commonality between redundant generators by way of the air supplies
or control systems.
2.3.9 Crank Case Breathers
The crank case breather is intended to allow oily fumes to escape from the engine crank case
and vent any slight pressure build-up. More severe pressure build-up related to crank case
explosion is vented by dedicated crank case doors or expanding bellows. Engine
manufacturers may recommend that crank case breathers are taken to the funnel top by way
of individual pipe work for each engine. In the case of vessels intended to work in explosive
atmospheres, the crank case breathers may be fitted with spark arresters.
Spurious operation of oil mist detectors can occur if crank case breathers become blocked,
and design problems related to fault tolerance and redundancy can arise if the pipe work for
crank case breathers is made common such that multiple engines may be affected by a
blockage in the pipe or spark arrester.
Engine room ventilation failures have been known to cause multiple engine loss when engine
safety shutdowns include crank case differential pressure. This occurs because the crank case
internal pressure is determined by the pressure outside the engine room through the crank
case breathers while the external pressure on the crank case is determined by the pressure
in the engine room. In DP Class 2 vessels with a single engine room this can cause loss of
engines exceeding the worst case failure design intent.
2.4 Engine Control and Safety Systems
2.4.1 Control Topologies
Most modern diesel engines have comprehensive and complex control and monitoring
systems. Various control topologies are possible using either the engine manufacturers
control system or the vessel automation system, as shown in Figure 22 and Figure 23
respectively. Various hybrid combinations which use the electronic speed governor as part of
the control system are also possible. Many DP vessel owners also specify unmanned
machinery space notations such as UMS, ACCU or E0, not because they intend to operate
with an unmanned engine room, but because these notations provide a suitable standard for
the control and safety systems required in a redundant propulsion system.
It is extremely important to understand the extent of control and safety functions provided
by the engine control system and the common mode failures that can be introduced by
connecting engines together through these systems. In general, redundant systems should
have as few common points between them as possible and any such common points should
fail to the safest condition and have a comprehensive set of protective functions designed to
prevent faults in one redundant element affecting another.

28 IMCA M 206
UPS
DIESEL ENGINE
ENGINE TERMINAL BOX
GOVERNOR ACTUATOR
TO VESSEL
SWITCHBOARD
ENGINE
SAFETY
SYSTEM
ENGINE
CONTROL
SYSTEM
AUXILIARY
SYSTEM
CONTROL
SPEED
CONTROL
ENGINE
CONTROLS
ECR
ENGINE
CONTROLS
WHEELHOUSE
VESSEL
AUTOMATION
PMS
ENGINE MANUFACTURERS
CONTROL SYSTEM
MAIN
POWER SUPPLY
FIELD
STATION
NET A NET B
EMERGENCY
POWER SUPPLY

Figure 22 Engine control and protection based on engine manufacturers systems
DIESEL ENGINE
ENGINE TERMINAL BOX
GOVERNOR ACTUATOR
TO VESSEL
SWITCHBOARD
ENGINE
SAFETY
SYSTEM
ENGINE
CONTROL
SYSTEM
AUXILIARY
SYSTEM
CONTROL
SWITCHBOARD
CONTROL
SPEED
CONTROL
UPS
VESSEL AUTOMATION SYSTEM FIELD STATION
NET A NET B
VESSEL AUTOMATION SYSTEM
LOCAL
CONTROL
PANEL
NET A NET B
VESSEL AUTOMATION SYSTEM
PMS AND
SWITCHBOARD
CONTROL FIELD
STATION
MAIN
POWER SUPPLY
EMERGENCY
POWER SUPPLY

Figure 23 Engine control and protection based on vessel automation system
2.4.2 Control Power
Engine control power is typically 24Vdc which may be derived either from a DC power
supply with battery backup or from a control system which is powered from a UPS.
Switchboard control supplies are typically 110Vdc and may mirror the 24Vdc distribution to
the engines. Occasionally, engine and switchboard control and protection functions are
supplied from a common DC distribution system. This arrangement introduces a common
point of failure between control and protection that can leave one or more uncontrolled
generators connected to the switchboard with no way to trip them. In such cases it may be
necessary to have another supply to trip the generators in the event that control power is
lost. The circuit breaker spring winder supply is sometimes used for this purpose.

IMCA M 206 29
The engine control and safety system should have separate power supplies. Ideally, these
supplies should be from separate sources, but class may accept separate fuses from the same
distribution. UPS battery endurance should be a minimum of 30 minutes. The UPSs (or DC
supplies) should have a dual power supply from the normal and emergency distribution. The
normal source of supply should be the main power system.
G3 G2 G1 G5 G4
DCA
G6
DCB

Figure 24 Typical control power arrangement
Figure 24 shows the most common supply arrangement where all engines on one bus share a
control power supply. However, there are significant benefits to making each generator as
independent as possible, even if the redundancy concept is only a two-way split. Losing
multiple generators as the result of a single failure is extremely disruptive to the power
system. Where the alternator has a permanent magnet generator consideration can be given
to using this as one source of control power such that each generator is independent of the
external control power source once it is running. Figure 25 shows just such an arrangement.
G3 G2 G1 G5 G4
DCA
G6
DCB
PMG PMG PMG PMG PMG PMG

Figure 25 Engine control power with PMG backup
Alternatively each generator can be supplied from its own control power system, as shown in
Figure 26. Where diodes are used to provide a dual source of supply, careful consideration
needs to be given to the possibility that an over voltage on one supply may damage all
equipment to which it is connected. A fault at the dual fed consumer may cause a voltage dip
on both supplies. Any redundancy concept must consider the effects of these types of faults.

30 IMCA M 206
G3 G2 G1 G5 G4
DCA
G6
DCB DCC DCD DCE DCF

Figure 26 Independent engine control supplies
In a few cases designers have elected to supply some engine control systems on each side of
the split in the redundancy concept from the same power source. Although this arrangement
ensures that a control power fault will not disable all generators on one bus, it may introduce
unacceptable restrictions on the combination of generators that can be connected as any
combination powered from the same source will result in loss of control over all running
engines, and possibly a blackout.
Cross connecting redundant engine control power supplies by way of diodes, as shown in
Figure 27, is also not recommended as a way of improving reliability as a voltage dip
associated with a fault in one control power supply will be seen by all control systems. If the
voltage dip ride through is insufficient then all running engines may malfunction. This
arrangement is also dependent on the selectivity of the fuses at the generator and power
supplies to ensure a fault in one engine supply does not blow the main fuses at each DC
supply output.
G3 G2 G1 G5 G4
DCA
G6
DCB

Figure 27 Diode isolated dual supplies
2.5 Safety Functions
2.5.1 General Requirements
Modern diesel engines are supplied with a large range of alarms and protection functions.
Specific requirements for safety systems depend on the size of the generators and also
whether the vessel has unmanned machinery space notation or not. The redundancy concept
needs to carefully consider the implications of each function which is capable of stopping the
engine. Some classification societies require that power failure in the safety system is not to
lead to a loss of propulsion. Engine safety systems are generally created using hardwired
relay logic. A few of the more common engine shutdown functions are discussed in the
sections which follow.

IMCA M 206 31
2.5.2 Over Speed
This may be a mechanical or electronic device and must be independent of the normal speed
control system. Sometimes the electronic over speed is based on a transducer sensing the
passing of teeth on the flywheel or similar arrangement. These devices normally fail safe
(engine continues to run) on failure of the power supply. However, there have been cases
where low voltage (rather than zero voltage) in the power supply to the transducer and
electronics has caused it to falsely indicate over speed and shut down the engine. In designs
where all engines share a common power supply it is possible for all engines to trip if the
power supply voltage goes out of tolerance.
Other control devices can also malfunction, certain models of governor are known to fail to
full fuel if their 24Vdc supply falls to 18Vdc.
2.5.3 Rig Savers
In the case of vessels operating in explosive atmospheres the over speed device must be
supplemented by a device that shuts off the combustion air supply, as discussed in Section
2.3.6. These devices normally fail as set on loss of control air supply.
2.5.4 Oil Mist Detection
The presence of oil mist in the crank case may indicate main bearing failure but also
represent a potentially explosive condition. In other cases the condition is caused by a
blocked crank case breather. Oil mist detectors are generally arranged to stop the engine on
detection of oil mist. The detectors generally operate on optical principles in which a sample
of the crank case gases is drawn through a sensing chamber using a Venturi effect. The air
supply to drive the Venturi is typically derived from the engine control air supply. Loss of the
air supply causes an instrument alarm but does not usually cause a false indication. Very
occasionally oil mist detectors on multiple engines have been known to operate spuriously in
response to certain tropical atmospheric conditions or if engines share crank case vents.
2.5.5 Crank Case Differential Pressure
Crank case differential pressure sensors are arranged to stop the engine if significant pressure
difference is detected. The presence of significant pressure in the crank case may indicate a
cracked piston, broken piston ring or water in the crank case. As it takes a short time for
the crank case and engine room pressure to equalise following a sudden change, loss of
engine room ventilation may trigger a spurious shut down on multiple engines. A time delay
may be introduced into the alarm and shut down circuit to overcome this problem.
2.5.6 Fuel Shut Down
Engine emergency stops and safety system may act in several ways to ensure an engine is
stopped:
governor signal to zero;
air cylinder pushes fuel rack to zero fuel a break back system may be used to
overcome the force of a faulty governor;
fuel solenoid valve will be closed;
electrically driven fuel pumps will be turned off.
2.5.7 JW Jacket Water Temperature High
High jacket water temperature may indicate a problem with the engine cooling water
systems. Loss of cooling water can cause engines and thrusters to overheat in a matter of
seconds, so operator intervention to isolate leaks is not a credible mitigation. It is for this
reason that some major classification societies require complete separation of freshwater
cooling systems for equipment intended to provide redundancy, even in DP Class 2 designs.

32 IMCA M 206
2.5.8 Emergency Stops
In addition to any other location required by class, engine emergency stops should be
provided for DP operators and engineers at the vessel management station close to the main
DP station. In DP Class 3 designs, line monitoring systems should be used to prevent loss of
multiple engines if emergency stop control lines pass through areas subject to fire or flooding.
2.6 Generators
2.6.1 Types of AC Machine
The generators used in the majority of modern DP vessels are three phase, synchronous, self
exciting, alternating current generators with brushless excitation systems. Variations on this
design includes:
self excitation based on build-up from remnant field;
permanent magnet exciter;
excitation supply from auxiliary stator winding.
In almost all self exciting generators the main DC field current is provided by way of shaft
mounted diodes from a small AC generator on the same shaft. This AC generator is called
the exciter. The stationary winding of the exciter is supplied with variable DC power from
the automatic voltage regulator (AVR) which may be supplied from another shaft mounted
AC generator called the PMG as shown in Figure 28, or by way of a transformer at the
generator terminals as shown in Figure 29. Some generators have an auxiliary stator winding
for this purpose, as shown in Figure 30. Self exciting alternators can excite without an
external source of power. In the case of generators that do require an external source of
power, class may require that a redundant source is provided for field flashing.
R
Y
B
POWER IN
CURRENT SENSE
VOLTAGE SENSE
POWER OUT
GENERATOR VT
AC
AC DC
DC AC
AVR
N
S
PMG EXCITER ROTATING
DIODES
ROTOR
FIELD
WINDING

Figure 28 Alternator with PMG

IMCA M 206 33
R
Y
B
CURRENT BOOST
POWER IN
CURRENT SENSING
VOLTAGE SENSORS
POWER OUT
AVR

Figure 29 Alternator with AVR powered by generator VT
R
Y
B
POWER IN
CURRENT SENSE
VOLTAGE SENSE
POWER OUT
AVR

Figure 30 Alternator with auxiliary winding
2.6.2 Short Circuit Performance
Classification societies typically require that alternators are able to deliver at least three times
rated current for 2s. Such large currents are often necessary to ensure the over current
protection scheme operates selectively. Alternators with permanent magnet generators or
auxiliary windings as the primary source of excitation power can maintain excitation through
the severe voltage dip associated with a close short circuit fault. However, alternators using
the terminal voltage as the source of excitation power require excitation support for the

34 IMCA M 206
duration of the short circuit fault. Excitation support is generally created by providing the
AVR with an alternative power supply from current transformers which allow it to derive
power from the fault current itself. Without excitation support, the generator may not be
able to deliver enough fault current to operate the over current relays and the voltage dip
may be extended to the point where the generator feeders or other circuit breakers trip on
under voltage protection, leading to blackout or widespread dislocation of the power
distribution system.
2.6.3 Generator Voltage Ratings
Generators for marine applications are typically rated for one of the many standard voltages
such as 380V, 440V, 480V, 690V, 3.3kV, 6.6kV and 11kV.
In marine rules and guidelines any voltage less than 1000V is considered to be low voltage
(LV) and anything above that is referred to as high voltage (HV). Shore-based utilities and
their equipment supplies may refer to marine HV levels as medium voltage (MV).
For any given power rating the line current falls as the generators voltage rating increases.
A significant factor in the choice of the system voltage level is the short circuit fault withstand
rating of the switchboards. Generally, the higher the fault withstand level the more expensive
the switchboard. Thus the higher cost of high voltage equipment may be offset by lower
short circuit fault withstand level. Higher current levels in LV equipment and distributions
also requires more copper and thus the cost, bulk and weight of cables may also have an
influence on the design. HV solutions tend to be favoured for power plants above about
10MW installed power.
Alternators may be fresh water cooled, air cooled or seawater cooled. Bearings may be
provided at both the non-drive end and the drive end. In some designs, the engine and
alternators share the drive end bearing.
Alternators are typically provided with a range of alarm and monitoring functions including:
current transformers for monitoring and protection;
stator winding temperature sensors;
air temperature indicators;
lubrication flow;
cooling water temperature indicators;
cooling water leak indicators.
The generator protection scheme is discussed in more detail in Section 2.12 on generator
protection philosophy.
2.7 Fuel Control
2.7.1 Compression Ignition
Diesel engines operate on the principal of compression ignition. Fuel is admitted to the
cylinders as the piston rises, compressing the air and increasing its temperature to the point
where it can ignite the fine spray of fuel.
There are three typical types of fuel control system for diesel engines:
Small multi-cylinder engines may have a rotary distributor type pump were each injector
is connected to a high pressure fuel pump in turn, according to the firing sequence;
Modern electronic diesel engines may have a common rail fuel system where a supply
of fuel is maintained at high pressure by the fuel pump and continuously distributed to
the injectors by high pressure pipe work. Injection is triggered electronically.
The amount of fuel admitted is controlled by the time each fuel valve is open;
Large medium speed diesel engines usually have one fuel pump and one fuel valve per
cylinder. The fuel pumps are operated by a camshaft and fuel is sent to the fuel valve

IMCA M 206 35
(injector) from the fuel pump at high pressure causing a spring loaded valve to lift from
its seat in the nozzle admitting fuel at the right time for combustion. The amount of fuel
admitted is controlled by varying the stroke of the fuel pumps. This function is carried
out by the fuel rack.
2.7.2 Engine Governors
The engine governor controls the fuel admission to the engine to maintain the desired speed
when the engine is operating independently and to maintain both the system frequency and
the desired share of the total system load when the generator is operating in parallel with
other generators.
There are various types of governors but the market for DP vessel generators is now
dominated by electronic governors with electric or electro hydraulic actuators to control the
fuel rack on the engine, as shown in Figure 31.
ALTERNATOR
ENGINE
CLOSE
OPEN
MAGNETIC PICK UP
SPEED
RAISE
SPEED
LOWER
DROOP/
ISOCHRONOUS
ACTUATOR SIGNAL
GENERATOR CURRENT
GENERATOR VOLTAGE
CB OPEN
CB CLOSE
CB STATUS
SPEED
POWER
24VDC
LOAD SHARING LINES
ACTUATOR
R Y B
GOVERNOR
FUEL RACK

Figure 31 Electronic governor with hydraulic actuator
Some governor actuators contain a mechanical ball head governor as backup to the electronic
unit. Although this feature was introduced to improve reliability it can be difficult to
coordinate with some load sharing systems and has caused blackouts in a few cases.
Many older vessels were fitted with hydro-mechanical governors. These units are often
interfaced to the power management system by a speeder motor (or pilot motor) which
physically adjusts the speed set point up and down in response to electric raise speed and
lower speed commands.
The governor is essentially a speed control system. The speed of the engines is monitored by
a magnetic pickup and transmitted to the speed control unit. Here the speed of the engine is
compared with the speed set point. The speed error is applied to a three term controller
which drives the fuel rack actuator to reduce the speed error. When two or more
generators are required to share load between them a power sensor is added to the control
scheme although it is possible to make generators load share without measuring the power
being delivered.
Electronic governors from the main manufacturers have reached such a level of sophistication
that they can be programmed to carry out many features normally found in vessel
management systems such as load sharing, synchronising, power management, alarm and
monitoring, and control of auxiliary systems such as fuel and cooling water pumps.

36 IMCA M 206
Fuel control for electronic engines is usually provided by the engine manufacturers
themselves and may be integrated into the engine control system as shown in Figure 32.
COOLANT
TEMP SENSOR
SHUTDOWNS
ELECTRONIC
GOVERNOR
(DESIRED RPM)
SIGNALS TO
FUEL
INJECTORS
TURBO OUTLET
TURBO INLET
TURBO ATM PRESS SENSORS
FUEL
INJECTION
CONTROL
TC #1
CYLINDER
SPEED SIGNAL
INTERPRETER
SPEED SENSOR
TORQUE
MAPS
ENGINE
CONTROL
LOGIC
(ENGINE RPM)
THROTTLE

Figure 32 Speed control for common rail electronic engine
2.7.3 Actuators
Governor actuators for electronic governors can be electric torque motors or electro-
hydraulic actuators. Actuators are described as forward acting if power is applied to drive
them towards the full fuel position and reverse acting if power is applied to drive them to the
zero fuel position, as shown in Figure 33. Forward acting actuators are the most common
but a few vessels were fitted with reverse acting actuators containing a mechanical ball head
backup governor. The principle of operations is that if the electronic governor fails, the
spring will pull the fuel rack to the operating point of the mechanical backup governor which
is adjusted to a slightly higher set point.
Forward acting actuators fail to the zero fuel position on loss of power or signal. This is a
relatively safe failure mode. Reverse acting actuators fail to the full fuel condition. This is not
a safe failure mode and can cause blackout unless another system intervenes. Electro-
hydraulic actuators derive their power from the engine by way of a gear drive from the
camshaft or other source. Electric actuators usually derive their power from a UPS or
battery charger supply associated with the electronic governor.
Problems can arise with actuators if they stick in one position.
ACTUATOR
FUEL RACK
FULL FUEL ZERO FUEL
GOVERNOR
REVERSE
ACTING
GOVERNOR
ACTUATOR
FUEL RACK
FULL FUEL ZERO FUEL
GOVERNOR
FORWARD
ACTING
GOVERNOR

Figure 33 Forward and reverse acting actuators
2.7.4 Speed Pickups
Engine speed is usually measured by a magnetic pickup sensing the passing of teeth on a
toothed wheel which may be the engine flywheel or mounted on an auxiliary shaft.
Sometimes, the toothed wheel is inside the actuator itself, mounted on the drive for the
hydraulic pump.
Failure of the magnetic pickup to no signal may cause the engine governor to go to full fuel
as it assumes the engine is not running fast enough. Most electronic governors have a

IMCA M 206 37
motoring function that will shut down the engine if the speed pick-up fails electrically. This
function can be turned off during certain maintenance procedures so it may be prudent to
confirm the function is active by periodically failing the speed pick-up when it is safe to do so.
There have been a few cases where the speed pick-up became loose and started to miss
pulses. This failure mode defeats the monitoring system and has caused a few blackouts.
Some governors are fitted with dual speed pick-ups. The governor will normally take the
highest reading pick-up as representing the engine speed. If one pick-up fails the governor
will issue an alarm but the engine will continue to operate. If the second pick-up also fails the
engine will stop. It may be necessary to stop and start the engine to clear the alarm on the
first pick-up after it has been repaired or replaced.
Dual speed pick-ups can also be used to improve frequency stability on generators with
flexible coupling between the engine and the generator. Such couplings have been known to
introduce very severe frequency oscillations leading to blackout if a single speed pick-up is
installed on the engine side of the coupling. Special software must be installed in the digital
governor to take advantage of this arrangement. If this feature is not available on the
governor in question, the pick-up should normally be installed on the generator side of any
flexible coupling to get the best results.
2.7.5 Load Sharing Schemes
There are four basic types of load sharing scheme:
speed droop;
compensated speed droop;
isochronous;
master slave.
2.7.6 Speed Droop
In this mode of load sharing, the governor is adjusted to allow the engine speed (system
frequency) to droop slightly as the load on the generator increases. Typical values for speed
droop are 3% to 5%. The generators speed may start to oscillate if the droop is too low and
2.5% is about the practical limit of stability. When two or more generators are operating in
parallel they naturally share load at a balance point where the common network frequency
intersects their respective droop lines. In Figure 34 the total system load is equal to the
rating of one generator. G1 and G2 are identical and have 3% droop represented by the
solid droop lines. Each generator therefore carries 50% of the available load at 60.9Hz. If the
droop line on G2 is offset upwards (dashed line) the system frequency rises to 61.25Hz and
load sharing becomes imbalanced with G1 carrying about 30% and G2 about 70%.
0 20 40 60 80 100
60
60.5
61
61.5
62
62.5
63
Load (%)
N
e
t
w
o
r
k

F
r
e
q
u
e
n
c
y

(
H
z
)
G1
0 20 40 60 80 100
60
60.5
61
61.5
62
62.5
63
G2
Load (%)

Figure 34 Generators load sharing by speed droop

38 IMCA M 206
2.7.7 Compensated Speed Droop
Power management systems use the method of offsetting the droop line in Figure 34 to
adjust both the network frequency and the relative load carried by each generator to
compensate for changes in total system load and engine/governor characteristics. In this
method of load sharing, which is sometimes called pseudo isochronous, the PMS is
constantly moving the droop lines on all the generator governors up and down to make each
generator carry an equal share of the load and maintain a constant system frequency across
the entire load range. To do this the PMS needs to know that a generator is connected and
what load it is delivering. The PMS also needs to know the bus frequency. Problems can
arise with this type of load sharing system in a number of ways:
A raise, lower signal sticks in the speed raise direction;
A generator circuit breaker gives false indication of status;
A generator kW transducer reads too low;
A bus frequency transducer reads too low.
Compensated droop can also be used to make one generator carry a much higher load for
engine conditioning purposes. This method of load sharing is often called asymmetric or
fixed target load sharing.
2.7.8 Isochronous
In this mode of control, the governors act to balance load sharing and maintain a constant
steady state frequency regardless of system load. The electronic control units for each
governor are directly connected by analogue or digital load sharing lines which transmit
information on generator load to the governors for all other connected generators. In the
case of analogue units, load bridges are used to develop a speed bias representing the
difference in the load carried by each generator, as shown in Figure 35. This bias is applied to
the speed control loop in each generator to balance load at the desired bus frequency.
MAIN
BUSTIE
LOAD SHARING LINES
POWER
DG 2
SPEED BIAS
POWER
DG 1
POWER
DG 4
POWER
DG 3
SPEED BIAS SPEED BIAS SPEED BIAS
GEN CB GEN CB GEN CB GEN CB

Figure 35 Analogue load sharing lines
In the case of digital load sharing Figure 36, the digital speed control units are connected by a
communications network.
DG 1
DSC
DG 2
DSC
DG 3
DSC
DG 4
DSC
COMMUNICATIONS
NETWORK
MAIN BUSTIE

Figure 36 Digital load sharing lines

IMCA M 206 39
Although synchronous generators run at exactly the same speed it is not possible to set the
speed set point of every generator exactly the same so the speed control loops will always
have a small error. This creates a very slight load sharing error that cancels out the speed
error, provided the load sharing control loop is working properly.
Problems can arise with this type of load sharing scheme if the load sharing lines fail. Some
manufacturers offer dual load sharing lines backed up by default to droop mode on loss of
communication or detection of significant load sharing imbalance. Although these features
mitigate many of the risks associated with control system failures they do not provide
comprehensive protection against a generator failing to full fuel and it is necessary to provide
a supervisory protection scheme to monitor generator performance independently of the
load sharing system and split the power system or trip a faulty generator before it can cause a
blackout.
2.7.9 Master Slave
Figure 37 shows a much simplified sketch of a master slave load sharing system. In this
arrangement one generator is always the master. There are two main parts to the control
system:
speed control;
load (real current) control.
The speed control loop acts to keep the master generator running at set point speed,
normally equivalent to 60Hz. The load controller then applies an additional governor
actuator command signal proportional to the load being carried by the generator, thus the
combined signal is intended to ensure the generator carries the applied load at the required
frequency. Only the master generator is under the control of its own speed control loop but
it supplies its speed error signal to all the slave generators for combination with the signal
from the real current sensor at each of the slave generator terminals. This system is not
widely used in modern DP vessels.
ENGINE
G
SPEED SET POINT
M S
M S
ENGINE
G
SPEED SET POINT
M S
M S
ENGINE
G
SPEED SET POINT
M S
M S
ENGINE
G
SPEED SET POINT
M S
M S
MASTER
SLAVE
SLAVE
SLAVE
REAL CURRENT
GOVERNOR
ACTUATOR
GOVERNOR
ACTUATOR
GOVERNOR
ACTUATOR
GOVERNOR
ACTUATOR
SPEED
REAL CURRENT
SPEED
REAL CURRENT
SPEED
REAL CURRENT
SPEED
S
P
E
E
D

E
R
R
O
R

O
F

M
A
S
T
E
R

G
E
N
E
R
A
T
O
R

Figure 37 Master slave load sharing system

40 IMCA M 206
2.7.10 Electronic Governor Principle of Operation
Figure 38 is a much simplified drawing of an electronic governor and its connections to a
generator. Most electronic governors are capable of operating in droop mode or
isochronous mode. The speed of the generator is monitored by the speed pick-up and
compared with the speed set-point to create a speed error signal to drive the actuator.
The speed error is modified by a signal from the real power transducer to create the
required speed droop as generator load increases.
The real power transducer measures the generators terminal voltage and line current and
calculates the three phase power being delivered in kW. The kW signal is applied to the load
bridge which has different functions depending on whether the governor is in droop mode or
isochronous mode.
In droop mode, one element of the bridge is imbalanced by adding R5 in parallel to create the
required droop signal. Unbalancing the bridge in this way causes the voltage difference (V
diff
)
signal to increase as generator load rises, as shown in Figure 39.
In isochronous operation, the load sharing lines are connected to the load bridges on all
other online generators at the point of circuit breaker closure. Any difference in the load
being carried by a generator will cause a current to circulate in the load sharing lines creating
a voltage difference on the output from the bridge (V
diff
). This signal then modifies the speed
error signal such that the signal to the actuator now admits fuel to correct for speed errors
and load sharing imbalance.
Effects of power sensor failures in droop mode: If the real power transducer fails to
no output, the generators speed and therefore the network frequency will rise to the no
load speed. This has the effect of unloading all the generators operating in parallel which also
run at their no load speed. If the total system load is greater than the rating of the faulty
generator, it may trip on overload in which case the load will be thrown back on the
generators operating in parallel. If the system load is less than the rating of the faulty
generator the load sharing imbalance may stabilise with the faulty generator carrying the total
system load and all other generators running at no load. However, if the no load set point of
the faulty generator is higher than the setting of the healthy generators there is a risk that
they will trip on reverse power. The exact outcome in this case depends on a number of
different factors, but blackout cannot be ruled out and protective functions should be
provided to split the power system or trip the faulty generator.
Effects of power sensor or load sharing line failure in isochronous mode: If the real
power sensor or load sharing lines fail in isochronous mode a similar effect occurs as there is
no signal to modify the speed set point and balance out the error between the actual bus
frequency and the speed set point. If the speed set point is lower than the actual bus
frequency then the integral part of the speed controller will continue to admit more and
more fuel in an attempt to reduce the speed to zero. However, this has the effect of
unloading the other generators to the point where they may trip on reverse power.
Experience of this fault in marine power system suggests the imbalance develops over a
period of five to ten minutes. A similar effect occurs if the load sharing lines break between
groups of generators.
Effects of power supply failure: Most types of electronic governors with forward acting
actuators fail to zero fuel if the supply voltage fails completely. However, certain models are
known to fail to the full fuel condition when the 24Vdc supply voltage falls to around 18Vdc.

IMCA M 206 41
ALTERNATOR
ENGINE
SPEED PICK UP
ACTUATOR
R
Y
B
FUEL RACK
SPEED
AMP -
-
+
PID
SPEED ERROR
OV
OV
24VDC
SPEED SETPOINT
SPEED
SPEED OFFSET (DROOP)
LOAD OFFSET (ISOCH)
R6
R5 R3
R4
R1
R2
V DIFF
CLOSED FOR DROOP
OPEN FOR ISOCH
CLOSE
OPEN
R
Y
B
24VDC
GENERATOR
CIRCUIT BREAKER
REAL POWER
TRANSDUCER
kW
ISOCH DROOP
LOAD BRIDGE
LOAD SHARING
LINES
V LOAD

Figure 38 Electronic governor

0 10 20 30 40 50 60 70 80 90 100
0
0.5
1
1.5
2
2.5
3
Generator Load (%)
S
p
e
e
d

O
f
f
s
e
t

(
%
)

Figure 39 Speed offset signal
2.7.11 Modern Digital Governors
Modern digital governors monitor several engine parameters to optimise engine efficiency
and reduced pollution. Vessel owners with identical engines having a mixture of analogue and
digital governors report superior engine performance from the numerically (digitally)
controlled engines. In particular, the load acceptance can be considerably improved.
Parameters monitored by digital governors include:
engine speed;
actuator travel;
turbocharger boost pressure;
jacket water temperature;
oil pressure;
engine load.

42 IMCA M 206
Figure 40 shows the block diagram of a typical modern digital governor.
MICRO PROCESSOR
ADDITIONAL
PROCESSOR
MEMORY
POWER
AMPLIFIER
CONVERTER
ACTUATOR
CONVERTER
MAGNETIC
PICKUP
SETPOINT
VALUES
BOOST
PRESSURE
SENSOR
TEMPERATURE
SENSOR
OIL PRESSURE
SENSOR
REDUNDANT ENGINE SPEED SIGNAL
ENGINE STOP
DIGITAL INPUTS
CONVERTER
CAN-BUS
ENGINE SPEED
ACTUATOR TRAVEL
ALARM
DIGITAL APPLICATION UNITS
e.g. SYNCHRONIZER &
LOAD MEASURING UNIT

Figure 40 Modern digital governor
2.7.12 Choice of Load Sharing System
The choice of speed control system will be influenced by many factors and each system has
advantages and disadvantages for each particular application. Load sharing systems based on
uncorrected governor droop introduce the fewest common points into a redundancy
concept and therefore offers advantages for DP Class 2 and DP Class 3 vessels intending to
operate a common power system. Operating the power plant in uncorrected droop mode
also allows the health of individual generators to be determined by monitoring active power,
reactive power, bus voltage and frequency.
Isochronous control systems maintain engine speed across the full load range and therefore
allow maximum power to be extracted from the generators. However, the commonality
introduced by load sharing lines and communications networks adds to the number of failure
modes that can destabilise load sharing and additional features, redundancy and protective
functions may be required to isolate faults in these connections.
Load sharing system using the PMS to compensate for governor droop add yet another layer
of control to the generators and thus introduce additional failure modes related to the
control interface between the governors and transducers used to measure power system
quantities such as power, voltage, current and frequency.
The choice of load control system is less important for vessels intending to operate the
power plant as two or more independent power systems as failure effects should be limited
to partial blackout. However, losing a power system while carrying out critical DP
operations always incurs some risk and designers may opt to apply the same level of
protection to each individual power system as they would to a common power system. The
benefit of this approach is that it reduces the failure effects of load sharing system faults to
loss of a single generator.
Thus, when specifying a load sharing system for a DP vessel there may be economic
advantages in including this function in the engine or PMS vendors scope of supply but any
advantage may be cancelled out if the need for special protective functions is not considered
at the same time.

IMCA M 206 43
2.8 Excitation Control
2.8.1 Excitation Systems
The various type of excitation systems have already been discussed in Section 2.6.1.
2.8.2 Automatic Voltage Regulators
The automatic voltage regulator is responsible for maintaining the generators terminal
voltage when the generator operates independently and maintaining the voltage and reactive
power sharing when the generator operates in parallel with others. Figure 41 shows a much
simplified schematic of a thyristor divert regulator that was popular on DP vessels.
CONTROLLER
FIRING PULSES
AC
DC
AC
DC
R
R
Y
B
DROOP ADJUST
FOR CROSS CURRENT
COMPENSATION
(OPTIONAL)
DROOP CT VOLTAGE SENSING FIELD SUPPLY EXCITATION CTs
N
FIELD
DIVERT
THYRISTOR
VOLTAGE
SET POINT
TO OTHER
GENERATORS
I DIVERT
I FIELD
+ -

Figure 41 Automatic voltage regulator (divert type)
The bridge rectifier has two sources of AC power. The normal source of supply is from the
red phase voltage by way of a choke in LV systems and a VT in HV systems. The alternative
source of supply is via the two excitation CTs on the red and blue phases, respectively. This
supply provides power when the terminal voltage is low as in the case of a close short circuit
fault. On the output from the DC bridge is the divert thyristor. This semiconductor is
controlled to divert current away from the generator field winding, thus controlling the
excitation level of the generator and thus the terminal voltage. Thyristor commutation
depends on the high inductance of the field winding which keeps the field current relatively
constant. The rectifier input current on the other hand is relatively sinusoidal due to the
impedance of the supply choke. When the rectifier current falls below the field current the
field current freewheels through the rectifier creating a forward volt drop, this reverses the
direction of the voltage across the thyristor turning it off.
The controller compares the generators R-Y line voltage against the voltage set point and
alters the firing angle of the thyristor accordingly. If the thyristor fails to the open circuit
condition, the field voltage current will go to maximum. Over excitation may cause the
operating point of other machines operating in parallel to move into the capacitive region and
trip on their field failure protection. If the thyristor fails to the on condition the faulty
generator will be tripped by its field failure protection.
2.8.3 Reactive Power Sharing Schemes
There are three common types of reactive power sharing schemes for automatic voltage
regulators. Most AVRs are capable of all three modes of control:
quadrature current compensation (reactive droop);
compensated reactive droop;
cross current compensation.

44 IMCA M 206
2.8.4 Quadrature Current Compensation
Quadrature current compensation (reactive droop) is added to the control scheme to allow
stable sharing of reactive power when generators are operated in parallel. It is analogous to
speed droop in governors and is by far the most common form of reactive power sharing
scheme for DP vessels.
In this form of reactive power sharing, the excitation level is allowed to fall slightly as the
amount of lagging reactive current increases. This is achieved by adding a small voltage
representing the reactive current being delivered to the sensed line voltage from the
generator VT. This makes the sensed voltage appear greater than it actually is, thus reducing
the excitation to create the voltage droop. A signal related to the reactive current is
obtained by measuring the blue line current with a CT. The current from the CT is used to
develop a voltage across a potentiometer. The voltage is then added to the red-yellow line
voltage before being applied to the controller as the sensed voltage.
Figure 42 shows that the blue phase voltage is 90 displaced from the red-yellow line voltage,
thus the blue line current will similarly be displaced by 90 plus however many degrees it lags
the blue phase voltage. By adding the signal derived from the blue line current CT (V
IB
) at 90
to the sensed voltage (V
RY
), as shown in Figure 43, the sensed voltage will increase and
decrease with the power factor seen by the generator. Although the sensed voltage will be
affected to some extent by both the active and reactive components of the blue line current,
the geometry is such that changes in the reactive component (in line with V
RY
) have much
more effect on the length of the sensed voltage vector than changes in the active component
which is at 90, as shown in Figure 44. Most marine loads are inductive and therefore the line
current lags the phase voltage creating the required droop as it increases. However, if the
line current is leading, the excitation level will increase as the current increases.
30
120
R
Y B
120
V
R
V
B
V
Y
60
V
RY
I
B

Figure 42 Relationship of phase and line quantities
V
RY
GENERATED
VOLTAGE
SENSED
VOLTAGE
VIB

Figure 43 Voltage representing blue line current is added to red-yellow line voltage

IMCA M 206 45
GENERATED
VOLTAGE
SENSED
VOLTAGE
UNITY pf
SENSED
VOLTAGE
LAGGING pf
ACTIVE
CURRENT
COMPONENT
REACTIVE
CURRENT
COMPONENT

Figure 44 Sense voltage increases as reactive component of current increases
2.8.5 Compensated Reactive Droop
In this form of reactive power sharing scheme the power management system adjusts the
voltage set points of AVRs operating in reactive droop in much the same way as it adjusts the
governor set points. As such, the PMS can trim the AVRs to maintain system voltage and
balance reactive power sharing.
2.8.6 Cross Current Compensation (Astatic Loop)
This form of reactive power sharing is analogous to isochronous speed control. Just as
isochronous control maintains the system frequency irrespective of load (no speed droop), so
cross current compensation maintains the system voltage irrespective of the amount of
reactive power being supplied (no voltage droop). Cross current compensation uses reactive
power sharing lines in much the same way that an isochronous load sharing scheme does.
This form of control is less common in marine applications.
Cross current compensation schemes makes use of the same droop CT and burden resistors
as used in the quadrature current compensation scheme with the addition of a loop
connecting generators operating in parallel, as shown in Figure 45.
CROSS CURRENT LOOP
AVR
G1
VL1
VR1
R1
DROOP CT
IP1
AVR
G2
VL2
VR2
R2
DROOP CT
IP2

Figure 45 Cross current compensation

46 IMCA M 206
The loop splits the current from the droop CTs with some passing through the burden
resistor as before and the rest circulating in the loop. When the reactive power being
delivered by the two generators is equal, the current circulating in the loop is equal and
opposite to that circulating through the droop CTs, with the effect that there is no voltage
across the burden resisters R1 and R2. If one generator delivers more reactive power than
the other for some reason, the current in the loop changes in such a way to create a positive
voltage across the generator with too much excitation and a negative voltage across the
generator with too little. This voltage difference across the resistors drives the two
generators back into balance again and the voltage across the resistors returns to zero.
So long as the voltage across the resistors is zero the sensed voltage will equal the actual
terminal voltage of the generators and the AVR will maintain the terminal volts at the set
point, regardless of the level of reactive power being delivered. Note that if the loop breaks,
both generators return to sharing reactive power in droop mode. If the line short circuits
then the resistors are effectively shorted out and one generator will take the entire reactive
load.
If the power system is using cross current compensation with more than two generators, it is
necessary to arrange for the loops to be created on either side of the busties so that the
generators on each independent power system can share reactive power. This further
complicates the switching arrangement required to break and terminate the cross current
loop.
V
R1
I
1
R
1
DROOP CT G1
V
R2
I
2
R
2
DROOP CT G2
R
1
= R
2
= 1
I
3
CROSS CURRENT LOOP

Figure 46 Cross current loop (voltage across resisters balances to zero)
2.8.7 Diode Failure Detector
The shaft mounted diodes that rectify the AC power from the exciter for use in the stator
field can fail. Some generator excitation systems have a diode failure detector which works
on the principle that rectifiers produce a characteristic ripple in their DC voltage waveform
and a faulty diode affects the frequency of this ripple. This effect can be used to activate an
alarm or shut down a faulty generator. Diode failure may eventually lead to excitation system
failure which can lead to blackout if suitable protection is not part of the generator
protection scheme.
2.8.8 Excitation Shutdown
Most AVRs or excitation systems will have some means of shutting down the generator field
in response to an external request. A typical application for this feature is the differential
protection used to detect a short circuit in the generators stator winding.
2.8.9 Digital AVRs
Analogue AVRs using the principles described above can be found on many DP vessels but
are increasingly being replaced by digital AVRs using numerical techniques to achieve the
same effects. These digital AVRs sometimes require a control power source independent of
the generator such as a 24Vdc battery/charger supply. The response of the AVR to failure of
this supply needs to be considered in the redundancy concept.

IMCA M 206 47
2.9 Main Switchboards and Motor Control Centres
2.9.1 Main Switchboards
Main switchboards can be rated for HV or LV use and are generally of metal-enclosed
construction. HV switchgear is generally designed to comply with standards such as IEC
60298, A.C. metal-enclosed switchgear and control gear for rated voltages above 1kV and
up to and including 52kV. Switchboards have a voltage rating, a continuous current rating
and a fault withstand current rating. The latter is intended to define the maximum fault
current the switchboard can physically stand without suffering damage. In some vessel design
the short circuit withstand rating limits the number of generators that can be connected to
the main switchboard at one time. When all generators are connected the main bustie opens
automatically to limit the fault current that can be experienced on each side.
Figure 47 shows a cross-section of a typical marine switchboard consisting of steel panels
arranged to carry a three-phase copper bus bar arrangement mounted on insulators within a
common enclosure, separated to divide the bus bar system from the cable compartments and
the circuit breaker. Air is used as the insulating medium but some bus bars may be insulated
along their length for additional protection. The circuit breaker can be withdrawn and is
interchangeable with those in other circuits. A low voltage enclosure is provided for
mounting instruments and control gear. Cable connects must be bottom entry for marine
applications.
A system of mechanical interlocks and shutters prevents access to live parts of the switchgear
when the circuit breaker is withdrawn and also prevents the carriage being withdrawn or
engaged when the circuit breaker is closed. Closing the circuit to the load using the primary
contacts of the circuit breaker when inserting the carriage with the circuit breaker closed can
result in severe damage.
Some classification societies require an arc proof rating for the switchboard and a pressure
relief duct to be provided to vent arc products safely out of the enclosure in the event of a
fault in the switchboard itself.

BUS BARS
WITHDRAWABLE
VACUUM CIRCUIT
BREAKER
CABLES
VOLTAGE
TRANSFORMER
EARTH SWITCH
CURRENT
TRANSFORMER
PRESSURE
RELIEF DUCT
INSTRUMENT AND
CONTROL CUBICLE
LOCAL
CONTROLS
PRIMARY
CONTACTS

Figure 47 Metal enclosed switchgear

48 IMCA M 206
2.9.2 Switchgear
Switchgear generally takes the form of circuit breakers or contactors. Circuit breakers for
marine applications are generally three pole and designed to close onto a short circuit fault
and open again without damage. Each circuit breaker has a continuous current rating and a
making and breaking capacity which indicate the fault current it is capable of interrupting.
Circuit breakers are generally capable of interrupting the current arcing across their open
contacts within three to five cycles of the power frequency waveform. Current limiting
circuit breakers are available for low voltage applications which operate quickly enough to
prevent the fault current reaching its peak level.
Power for the circuit breakers closing mechanism is derived from a motor wound spring.
The powerful spring which provides the force to open the contacts is charged by the action
of closing the circuit breaker. The spring mechanism is recharged by the motorised spring
winder every time the circuit breaker is closed. Enough energy is stored in the mechanism
for an open-close-open cycle. Circuit breakers can have different types of insulating medium
including:
air;
vacuum;
SF
6
;
oil.
Very few marine applications make use of oil filled circuit breakers but they found limited
application in fixed offshore structures.
Contactors are electromagnetically operated switching devices. An AC or DC coil is used to
pull the switching contacts together against the force of a spring. Loss of current to the coil
will cause the contactor to open again. Some designs are mechanically latching, in this design,
a pulse of current is sent to close the contactor and another to open it. This type of
contactor remains as set on loss of control power. Contactors are only designed to handle
load current not fault current and thus contactors must be paired with a suitable circuit
breaker or fuses.
2.9.3 Motor Control Centres
Motor control centres (MCCs) contain groups of motor starters and may form a significant
part of the low voltage distribution in a marine power system. A motor starter for an LV
circuit typically contains a moulded case circuit breaker (MCCB), a contactor with thermal or
magnetic overload protection and relay logic to allow remote starting and stopping of the
motor. Modern motor control centres may be controlled by a dedicated communications
network or by hardwired contacts to a local vessel management field station.
2.10 Power System Faults
2.10.1 Critical Power System Parameters
A power system can be said to be in a fault condition if any of its critical parameters are out
of tolerance for more than an acceptable time period, such as that associated with expected
power system transients.
Quantities which must remain within tolerance include:
voltage;
current;
frequency;
levels of harmonic distortion;
line current balance;
phase voltage balance.

IMCA M 206 49
2.10.2 Fault Conditions on Three Phase System
Short circuit one or more phases;
open circuit one or more conductors;
earth fault;
over/under frequency;
over/under voltage;
over load rating of engine exceeded;
over current rating of alternator, bus bar, cable, motor, transformer or other
consumer exceeded;
severe active power sharing imbalance;
severe reactive power sharing imbalance;
excessive regeneration of power;
severe waveform distortion;
loss of synchronisation and crash synchronisation.
2.11 Overall Protection Philosophy
2.11.1 Purpose of Electrical Protection Schemes
Electrical protection schemes are designed to prevent the uncontrolled release of energy
associated with power system faults, thus protecting life and limiting damage to equipment.
In the case of DP vessels of DP Class 2 and DP Class 3 the power system protection scheme
must also ensure continuity of supply to essential consumers such as thrusters and auxiliary
systems. The protection scheme must be coordinated to ensure that faults are isolated as
close to source as possible and that failure effects do not exceed the worst case failure design
intent. The primary protection function in any electrical protection scheme is over current
protection, which is intended to prevent excessively high currents causing cables to catch fire.
The overall electrical protection scheme can be broadly divided into the following sections:
generator protection;
bus bar protection;
feeder protection.
Generator protection is provided to limit the effects of internal faults in the generator, to
protect the generator from the effects of power system faults and protect the power system
from the effects of generator faults. Generator protection is discussed in more detail in
Section 2.12.
2.11.2 Selectivity
The term selectivity can be applied to the ability of any protective function to isolate a fault as
close to source as possible, but it is most often applied to the selection of fuses or the setting
of protection relays for circuit breakers to ensure that faults are disconnected by the circuit
breaker or fuse closest to the fault. The purpose of these protection devices is to detect
faults and isolate them quickly and selectively from the system therefore limiting the
consequences of the fault as far as possible.

50 IMCA M 206
10
2
10
3
10
4
10
5
10
-2
10
-1
10
0
10
1
10
2
Current (A)
T
i
m
e

(
s
)
Normal Inverse Time-Current Characteristic
10
2
10
3
10
4
10
5
10
-2
10
-1
10
0
10
1
10
2
Current (A)
T
i
m
e

(
s
)
Definite Time Current-Characteristic

Figure 48 Definite and inverse time-current characteristics
Selectivity can be described as full or partial. For full selectivity:
the circuit breaker must not trip in response to current transients associated with
normal operation;
the circuit breaker closest to the fault should trip before any upstream circuit breaker;
if the circuit breaker closest to the fault fails to trip, the upstream circuit breaker should
operate to clear the fault.
In the case of partial selectivity, the circuit breaker closest to the fault will only trip first up to
a certain current level. If that level is exceed the upstream circuit breaker may operate first.
There are essentially two approaches to over current protection, definite time grading and
inverse time grading ,examples of these characteristics are shown in Figure 48. In the case of
the definite time characteristic the tripping time is constant regardless of the fault current
level. Conversely, the tripping time of the inverse time relay reduces as the fault current
increases. There are various forms of the inverse time current characteristic, including:
normal inverse;
very inverse;
extremely inverse;
long time inverse;
inverse definite minimum time.
Equations for these characteristics are defined by IEC and ANSI.
The inverse definite minimum time curve has an inverse characteristic for the upper part and
a definite time for the lower part.
Figure 49 shows how a two-stage definite time over current relay may be used to prevent
blackout in the case of two switchboards connected by a bustie.
Figure 50 shows the time-current characteristics for the generator and bustie circuit breaker
over current relays. The upper part of each characteristic is called the low set over current.
In the case of the generator this is set slightly above its current rating. The lower part is
called the high set over current and is set to ensure the bustie opens first for short circuit
faults.
Because the characteristic of the bustie circuit breaker is separated from that of the
generator circuit breaker in both time and current the bustie should open first for all fault
currents up to the maximum possible fault current. If there were no selectivity between the

IMCA M 206 51
generator and bustie circuit breaker the generator on the healthy switchboard may trip first
and the vessel would blackout out.
In very general terms, two circuit breakers will be selective with respect to each other if the
time-current characteristic of the circuit breaker closest to the fault is to the left of and
below the characteristic of the upstream circuit breaker. However, the characteristics may
cross over at the low set over current to allow several generators to feed power across a
bustie to loads on the other switchboard, in this case the bustie should be rated to carry
these current levels. Other forms of bus bar protection are described in Section 2.11.4.
DG2 DG1
FAULT
BUSTIE
I>
I>>
I>
I>>
I>
I>>

Figure 49 Switchboards connected by bustie with over current relay
10
1
10
2
10
3
10
4
10
5
10
6
10
-2
10
-1
10
0
10
1
10
2
10
3
10
4
10
5
Current (A)
T
i
m
e

(
s
)
Coordination of Generator and Bustie Circuit Breakers
Bustie
Generator

I>>
I>
Short circuit current

Figure 50 Co-ordination of generator and bustie circuit breakers
In general terms, short circuit faults should be disconnected very rapidly to prevent damage
but low level over current conditions can be tolerated for a longer time and time-current
characteristics can be combined to achieve the desired response.
The study carried out to determine the shape of characteristics to use and the associated
relay settings is often referred to as a selectivity study. The terms discrimination study and
co-ordination study are also used.

52 IMCA M 206
Circuit breaker and protection device manufacturers may place certain restrictions on
settings. Selectivity may not be possible outside these limits.
The example given in Figure 50 is only intended to illustrate the basic theory and the
protection arrangements of large modern DP vessels can be very complex by comparison.
Protection engineering is a specialist branch of electrical engineering and selectivity studies
are now routinely carried out using computer programs.
2.11.3 Protection Relays
Modern protection schemes use digital protection relays to detect a large range of fault
conditions and trip the appropriate circuit breakers to isolate the fault. The relays are usually
mounted in the instrument and control cubicle of the switchboard and make measurements
of the power system voltage and current from voltage and current transformers (VTs and
CTs) connected to the bus bars and cable ways. Current transformers for protection duties
require a special rating which indicates their ability to measure fault currents without
saturation. Modern multi-function relays can be programmed for many different protection
tasks and can also be used as transducers to feed power system information to other systems
for monitoring and display. However, care should be taken not to link control and
protection functions in such a way that protection and control functions can be lost at the
same time. Failure to separate control and protection functions may lead to a situation
where a single failure causes machinery to adopt a dangerous condition and render the
automatic protection inoperative at the same time.
Protection relays are programmed with the settings from the protection co-ordination study.
Programming may be carried out by way of the front panel or by attaching a dedicated
programming tool or laptop. Protection relays typically require 110Vdc power for operation,
which is supplied from a dedicated charger rectifier with battery bank.
ANSI device codes for protection relays are widely used. Some of the more common codes
are listed below:
50 instantaneous over current relay;
51 AC inverse time over current relay;
64 ground (earth) detector relay;
27 under voltage relay;
40 field (over/under excitation) relay;
59 over voltage relay;
60 voltage or current balance relay;
67 AC directional over current relay;
86 lockout relay;
87 differential protective relay;
81 frequency relay;
32 directional power relay;
25 synchronizing or synchronism-check device.
2.11.4 Bus Bar Protection
Bus bar faults are the least likely failure in a marine power system but a great deal of time and
cost is expended trying to ensure that a fault acting directly on the bus bars does not cause
the worst case failure design intent to be exceeded. In fact, bus bar faults are so unlikely that
some vessel owners have successfully negotiated an exemption from class for bus bar faults in
relation to determining the vessels worst case failure and, therefore, its post-failure DP
capability.
Bus bar protection is designed to isolate the effects of short circuit and earth faults acting
directly on the bus bars or their connections. Bus bar protection can take the following
forms, depending on the number of bus sections that have to be protected:

IMCA M 206 53
over current protection;
differential protection;
directional over current protection;
optical arc detection;
pressure detection;
earth fault protection.
2.11.5 Over Current Protection Quick Trip
Figure 51 shows the most common arrangement for vessels with a simple two-way split.
In the event of a short circuit on either bus all connected generators will feed the fault. Each
generators over current relay can trip the generator and one of the main busties. Time
delays are arranged so that the busties open after 0.3s isolating the fault to one bus or the
other. The generators on the healthy bus section will no longer see the fault and their over
current relays will reset. Those on the faulty section will trip after 1.0s, isolating the fault
(lower tripping times may be possible in practice). A fault in a feeder is cleared by dedicated
feeder over current protection within the 0.3s time delay. A fault on a generator is cleared
near instantaneously by the generators differential protection. Thus the protection scheme is
fully selective for faults in any part of the power system. Some classification societies are
considering introducing a requirement to trip the busties first for all faults no matter where
they are in the system. Some vessel power plant designers have extended the quick trip
principle to cover many other faults that might destabilise the power plant, including but not
limited to:
over/under voltage;
over/under frequency;
over current;
current imbalance.
OVERCURRENT
0.3s
G3 G2 G1 G4
OVERCURRENT
1.0s
NC
OVERCURRENT
1.0s
OVERCURRENT
0.3s
OVERCURRENT
0.3s
OVERCURRENT
1.0s
NC
OVERCURRENT
1.0s
OVERCURRENT
0.3s

Figure 51 Time graded over current protection
2.11.6 Differential Protection

Figure 52 shows one of the popular solutions for bus bar protection in multi-split power
distribution systems. The challenge for the protection scheme in a multi-split system is to
isolate only the faulty bus section leaving all others connected together. This is particularly
important if some bus sections have no generator connected. Differential protection works
on the principle of Kirchhoffs current law which states that the sum of the currents entering
a node must equal the sum of the currents leaving the node. The bus bars are considered to

54 IMCA M 206
be the node and the currents being supplied or consumed by all the generators, feeders and
busties are summed by the protection relay for that zone (magnitude and phase angle).
The boundary created by the location of the current transformers used to measure the
currents flowing in and out of the bus section is called the protection zone. For this reason
differential protection is sometimes known as zone protection. If the bus bar is healthy the
current measurements will sum to zero. If however, the bus bar is faulty there will be a
current path within the zone which is not balanced out. This imbalance is detected by the
protection relays and the bus section is isolated by tripping the busties at the end of the zone.
Tripping times for zone protection are relatively fast and generally less than 150ms. In some
schemes the generators in the faulty zone may also be tripped by this function. To ensure
complete coverage it is common practice to overlap the CTs for each zone. This does
create a very small area common to two zones. A fault exactly at this point could cause two
zones to trip exceeding the worst case failure design intent. Additional complexity can be
added to overcome this but this may not be justified by the risk.
G3 G2 G1 G5 G4
NC
G6
NC
NC NC
D
C
NC NC
A
B

Figure 52 Differential protection
Faults at A B C and D have the following effects:
A fault at point A causes the LHS bus section to be isolated by the zone protection.
The centre and RHS bus sections remain connected;
A fault at point B will cause the tie line to trip leaving all other bus sections connected;
A fault at point C will be detected by the generators differential protection. Only G4
will be tripped;
A fault at point D will be detected by the thrusters over current protection. Only the
thruster will be tripped.
The number of CTs required to implement the scheme makes it more costly than the other
solutions but it does allow the power system to be run as a closed ring and therefore allows
complete flexibility in generator utilisation. There have been problems with spurious tripping
of zones in response to large motor or transformer starting transients and the efficacy of
differential protection schemes is related to the quality of the protection equipment specified.
2.11.7 Directional Over Current Protection
Directional over current protection offers a slightly cheaper alternative with many, but not
all, of the advantages of differential protection. Directional protection schemes are generally
operated in split ring configuration to establish well defined fault current paths so there is no
ambiguity about which circuit breakers should be blocked and which should be tripped, Figure
53 illustrates the general principle. Directional over current relays at each bustie are
arranged to block the circuit breaker up stream of the fault from tripping, thus only the

IMCA M 206 55
circuit breakers closest to the fault trip. Because the ring is split there may be no current
through some bustie circuit breakers. Generators left connected to the faulty bus section
will trip on over current. In Figure 53 a fault at point A will cause the bustie circuit breaker
at the left hand end of the G5, G6 bus section to trip. All upstream circuit breakers which
see fault current are blocked.
G3 G2 G1 G5 G4
NO
G6
NC
ONLINE ONLINE ONLINE ONLINE ONLINE ONLINE
NC NC NC NC
BLOCKED BLOCKED BLOCKED
A
TRIPPED

Figure 53 Directional over current protection
Figure 54 illustrates one of the disadvantages of directional protection which is that if one of
the bus sections has no generators connected, then two bus sections will be lost if a fault
occurs at point A. Thus the vessel may lose more thrusters than desired. In some
arrangements it may be possible to overcome this disadvantage by always running at least one
generator on each bus or by changing the point at which the ring is split to ensure there are
always generators at the extremities of the circuit. This would need careful coordination
with the power management system running order selection.
G3 G2 G1 G5 G4
NO
G6
NC
OFFLINE OFFLINE ONLINE ONLINE ONLINE ONLINE
BLOCKED
A
TRIPPED
NC NC NC NC

Figure 54 Directional over current protection
2.11.8 Optical and Pressure Arc Detection
This method of bus bar protection is based on an entirely different principle. Faults in
switchgear are generally associated with arcing faults. Thus it is possible to create a
protection system which will rapidly trip the bustie circuit breakers for the faulty bus section
if the light from an arc is detected inside the bus bar chamber. The main purpose of this type
of protection is to decrease the disconnection time to reduce the damage, hazard and repair
time. Disconnection times are very short, of the order of 100ms, and do not affect the
discrimination of other protection. Commercial systems use fibre optic cables to monitor
the internal spaces of the switchboard. Systems based on detecting the pressure rise
associated with arc products work on a similar principle. Some vessel power plant designers
choose to use arc detection as the primary protection scheme which supports the vessels
worst case failure design intent but provide more conventional over current protection as a
backup. The over current protection is not sufficiently selective to prevent WCFDI being
exceeded but it should prevent a fire if the optical system failed for any reason.
2.11.9 Earth Fault Protection and System Earthing
The type of earth fault protection specified for marine systems is influenced to some extent
by the size of the power distribution system and the maximum prospective earth fault
current. Many LV marine power systems were designed as isolated systems or not

56 IMCA M 206
intentionally earthed systems, where the power system has no direct connection or
reference to earth (vessels hull). On these systems, earth faults were typically indicated by
earth fault lamps connected from each line to earth. When one lamp goes dark and the
other two go bright there is an earth fault on the line connected to the dark lamp. Earth
insulation meters are the modern alternative. These devices can be connected to the alarm
and monitoring system.
On high voltage systems, high resistance earthing of various types is generally employed. In
these systems a high impedance path for earth fault currents is created by adding a resistor
from the generator star point to the ships hull or by way of one of several types of neutral
earthing transformer that can be connected to the bus, as shown in Figure 55. Earthing the
power system, by way of the generator star points, causes the earth fault current to vary
with the number of generators connected. Nevertheless, this continues to be a popular
method of system earthing.
Modelling a power systems response to earth faults can provide useful information in relation
to earth fault current and also the voltage transients that can be associated with such faults.
Such studies may allow the earth fault protection to be designed to operate more effectively.

R
Y
B
SYSTEM
CAPACITANCE
GENERATOR
+ + +
+ + +
OPEN DELTA
NEUTRAL
EARTHING
TRANSFORMER
ZIG-ZAG
NEUTRAL
EARTHING
TRANSFORMER
NEATRAL
EARTHING
RESISTOR
VCB

Figure 55 Methods of marine power system earthing
All power systems are referenced to earth by way of the distributed capacitance of cables and
windings and a significant earth fault current can flow even in not intentionally earthed
systems. The intentional earth impedance only adds to the system charging current when an
earth fault occurs and is often sized to provide an earth fault current three times that which
would flow as a result of the capacitive charging current. This provides well defined current
paths for protection purposes.
Classification societies differ on the level of earth fault current that can be accepted without
automatic isolation but the level is generally of the order of a few amps to a few tens of amps
and may be derived from the definition of high resistance earthing in some cases.
Some vessels in service only have an alarm to indicate an earth fault on the HV power
system, others have fully selective earth fault protection schemes. The requirement for the
voltage rating of HV cables differs for systems intended to operate continuously with an earth
fault and those with automatic disconnection. IEC standard 60092-352, (third edition),
Electrical Installations in Ships Choice and Installation of Cables, defines three categories of
power system: A, B and C. The categories relate to the amount of time the system can be
expected to run with an earth fault. Category A systems must have automatic disconnection
of an earth fault. Category B systems accept that an earth fault could be present for up to
eight hours at a time. Category C covers systems not covered by A or B.

IMCA M 206 57
The choice of cable to use in each case is based on standard cable voltage ratings using the
designation U
0
/ U / (U
m
) where:
U
0
= rated voltage between conductor and earth (screen);
U = rated voltage between conductors;
U
m
= maximum sustained system voltage.
The term U
m
is chosen to be equal to or greater than the highest system voltage. In
circumstances where a cable is specified for use on a system where the voltage exceeds the
nominal cable ratings it shall not exceed U
m
.
U
0
(kV) U (kV) U
m
(kV)
3.6 6.0 7.2
6.0 10.0 12.0
8.7 15.0 17.5
12.0 20.0 24.0
Table 5 Typical IEC voltage ratings for cables
On high resistance earthed power systems or isolated power systems of category A or B, a
cable would be chosen to match the rated system phase and line voltages. For example, on a
6.0kV power system the cable chosen would be rated at 3.6/6.0/(7.2kV).
In the case of category C, where the earth fault may exist for longer than eight hours, a 6.0kV
power system would require 6.0/10.0/(12kV) cable. This is necessary because the line to
earth voltage increases by 3 when an earth fault occurs.
Some classification societies only define two categories:
i) with automatic disconnection;
ii) without automatic disconnection.
In the case of a power system with automatic disconnection of earth faults a cable is chosen
to match or exceed the nominal system voltage. If there is no automatic disconnection a
cable must be chosen with a U
0
rating equal to or greater than the nominal conductor to
conductor voltage.
The discussion above is only intended to illustrate the process. Readers are advised to
consult the rules and standards directly before choosing a cable rating for any particular
application.
As the use of automatic disconnection of earth faults allows the power system designer to
use cables with a lower voltage rating there may be some commercial advantage in specifying
this type of protection. This advantage may be offset by the cost of an earth fault protection
system capable of selectively isolating an earth fault in a generator, bus bar or feeder. In the
worst cases, a crude earth fault protection scheme may be specified which simply trips all
generators on detection of an earth fault anywhere on the system. With such an
arrangement the power system must be operated as two or more independent power
systems to remove the risk of total blackout following an earth fault.
Earth fault protection for the main power system is sometimes based solely on time grading.
The relay in the earthing resistor or earthing transformers for each bus will detect an earth
fault at any point in the plant not isolated by a transformer.
Earth fault protection in the feeders will isolate a fault in a consumer. If the earth fault
persists after the tripping time of the feeder the fault is assumed to be in the generators or
on the bus bars itself. At this point the protection driven from the neutral earthing
transformers will trip the main busties to limit the earth fault to one bus or the other.
Whichever neutral earthing transformer continues to detect an earth fault will then trip all
generators connected to that bus. As losing a whole bus because of an earth fault in one
generator is unnecessarily severe some designers add restricted earth fault protection to the
generators.

58 IMCA M 206
Restricted and unrestricted earth fault protection principles are widely utilised with a
selection of protection relays to provide selectivity. Typical earth fault protection relays have
IDMT, DTL and direction elements. Core balance measuring (where the sum of any three
phased balance or unbalanced loads equals zero) is a common method of detection which
achieves selectivity when combined with time settings.
2.11.10 Feeder Protection
Feeder protection is generally limited to over current and earth fault protection but specialist
protection functions may be used to protect motors and transformers. In particular,
protection functions for transformers may need to be desensitised to the inrush current
transient that occurs when a large transformer is connected. In protection schemes where
the generators are protected against the effects of unbalanced currents (sometimes referred
to as negative phase sequence protection (NPS)) it may be prudent to include NPS protection
in the feeder protection scheme and co-ordinate it with the NPS protection in the generators
to prevent a large unbalance current originating in the distribution scheme causing all online
generators to trip. Such faults are uncommon in marine power system but cannot be ruled
out completely. NPS can also be used to trip the main busties as additional protection
although this may only be suitable for power systems with a simple two-way split.
Feeder circuit breakers for service transformers may be fitted with under voltage release to
disconnect transformers on blackout. It is important that an under voltage trips have
adequate time delays to prevent the trip operating on a voltage dip associated with a short
circuit fault. If this is not the case, all marine auxiliary services may stop due to an unrelated
fault somewhere in the power system.
Overload relays are generally used on equipment that can overheat such as generators,
motors and transformers. The temperature conditions of the equipment are simulated with
the same time constant in the protection relay. The heating and cooling curves used to
create the thermal model within the relay are selectable.
2.11.11 Power System Studies
Several studies may be commissioned to support the design of a marine power systems
including:
short circuit calculations;
protection co-ordination study;
load balance;
harmonic analysis;
transient stability study.
Short circuit calculations are performed to ensure the switchgear is able to withstand the
forces generated by the worst case short circuit current. It is also used to ensure the circuit
breakers are able to interrupt that level of fault current. Studies are carried out using
computer programs based on a recognised standard such as IEC 61363-1 Electrical Installations
of ships and mobile and fixed offshore units, Part 1: Procedures for calculating short-circuit currents in
three phase a.c. When calculating the contribution to short circuit current it may be
necessary to consider the contribution from motors and certain types of drives in addition to
the fault current delivered by the generators.
The protection co-ordination study (sometimes known as the discrimination or
selectivity study) is carried out to determine the various protection settings necessary to
ensure that faults are isolated as close to source as possible.
The load balance is used to show the power consumed under various operating
conditions, which may included DP, transit and harbour with variations for summer and
winter operation if appropriate.
The harmonic analysis is used to show that levels of harmonic distortion fall within
acceptable levels under all expected operating conditions. Excessively high levels of harmonic

IMCA M 206 59
distortion have been known to cause equipment malfunction exceeding worst case failure
design intent.
Transient and voltage stability studies identify the ability of the generators in a power
system to maintain synchronism when subjected to a severe transient disturbance such as a
fault, sudden loss of generating capacity or large load rejection. They can also be used to
ensure that motors can restart and that generators can restore voltage. These studies are
not always carried out on marine power systems because of their compact nature but some
of the more unusual designs do have additional impedance between generators.
2.12 Generator Protection Philosophy
2.12.1 Importance to Redundancy Concept
The range of protective functions applied to the generators is an important consideration in
the design of any DP vessel power plant. However, in the case of a vessel intending to
operate with a common power system, the generator and bus bar protection is fundamental
to ensuring fault tolerance and the integrity of the redundancy concept. It is important to
understand the dual role of generator protection in a DP vessel application. The protection
should provide the necessary level of safety but also ensure continuity of supply and limit the
severity of the failure effect to within the worst case failure design intent. Adding
inappropriate protective functions can be as damaging to the redundancy concept as having
insufficient protective functions.
GENERATOR PROTECTION RELAY
GROUND
CT
OPEN
RTDs STATOR
CB
STATUS
VCB
R
Y
B
SERIAL LINK
TO ALARM
AND
MONITORING
ALTERNATOR
V MW MVAR H z
4-20 mA
POWER
110Vdc
TO BUS TIE
CONTROL CCT
TRIP
BUSTIE
TRIP
GEN CB
FROM
AVR
FROM
GOVERNOR
WITH TRIP
COIL
SUPERVISION
ACTUATOR
CURRENT
FIELD
CURRENT
VTs CTs CTs
DIGITAL INPUTS
EXTERNAL
PROTECTION
CB TRIP
EXTERNAL
PROTECTION
TIE TRIP
PMS
OPEN
PMS
E STOP
START
ALL
DIESELS
RELAY
FAILURE
COMMON
ALARM
EXCITE
TRIP
TO PMS

Figure 56 Generator protection relay
2.12.2 Standard Generator Protection Functions
Figure 56 shows a typical multifunction generator protection relay. All information regarding
the health of a generator is obtained from VTs, CTs and winding temperature sensors. Table
6 shows a typical list of protective functions that might be available within such a
multifunction relay and the executive action that is taken on detection of each type of fault. It
is important to note that this range of protective functions is not sufficient to ensure the fault
tolerance of a common power system. Section 2.13 on advanced generator protection
discusses the additional protection features required for fault tolerance.
From Table 6 it can be seen this protection relay can be programmed to:
trip the generator;
trip the bustie circuit breaker between the two main switchboards;

60 IMCA M 206
shut down the faulty generators excitation system;
signal the PMS to start another generator;
lockout the generator from reconnection;
activate an alarm.
Function
Generator
Trip
Bustie
(Quick
Trip)
Excitation
Trip
Start All
Generators
Gen
Lockout Alarm
Phase differential

Negative sequence

Under-voltage

Reactive power

Over-voltage

Phase reversal

Under frequency

Loss of excitation

Reverse power

Phase over-current

Over frequency

IAS/PMS E-Stop

IAS PMS CB open

High set over-current

Trip coil monitor

VT fuse failure

Diode failure

Gen winding RTD

Gen bearing RTD

Field current

Relay fault

ESD

Earth fault

Table 6 Generator protective functions
2.12.3 Phase Current Differential
This function uses current transformers located at both ends of the generator windings. Any
fault path within the generator is seen as an imbalance by the protection relay. The function
is only required on larger generators (typically 1500kVA and above) and is designed to detect
stator winding faults at a lower level of fault current than the phase over-current function.
It will trip the affected generator very rapidly on detection of this fault.
2.12.4 Negative Phase Sequence
This function is designed to protect the generator from damage associated with a significant
imbalance in the generator phase currents. A phase imbalance creates a backwards rotating
field in the machine which may cause the generator to overheat. A broken conductor on a
generator, drive transformer or service transformer feeder could have this effect. Such
failures are very unlikely but if the overall protection scheme is not fully selective then all
generators which see the fault may trip leading to blackout. Certain types of drives may draw
unbalanced currents in certain failure modes without necessarily tripping immediately. It may
be possible to configure this function as an alarm rather than a trip with the agreement of the
generator manufacturer.

IMCA M 206 61
2.12.5 Under Voltage
This function is commonly applied to generators and is a class requirement. It has no
particular benefit for the redundancy concept except to ensure that the plant can be
restarted after a blackout. One possible way in which an under voltage could occur is the
voltage dip associated with short circuit faults. These faults should be cleared by the over-
current scheme and the under voltage protection should have a suitable delay to prevent it
reacting to this. Care must be taken when determining the settings of under voltage relays as
all connected generators operating in parallel will experience the same voltage and frequency.
Under voltage could also be caused by the direct online starting of a large motor.
2.12.6 Reactive Power
There are considered to be two ways in which generators could be called upon to supply too
much reactive power to the system. The first is due to an excitation system fault in one
generator. This fault affects the operation of generators running in parallel with the faulty
machine. The second is for the reactive power demand of the drilling or propulsion system
to exceed the capacity of the online generators. In general there need to be other protective
functions in place to prevent the power plant reaching this condition.
2.12.7 Over Voltage
This condition might occur as the result of an AVR failure to full or over excitation. This
function has very little benefit for the redundancy concept as it is not fully selective and
cannot identify the source of the over voltage and disconnect it. Other protective functions
need to be in place to prevent the power plant reaching this condition.
2.12.8 Phase Reversal
This fault is unlikely to be present beyond the commissioning phase but could occur after
repair.
2.12.9 Under Frequency
This is a symptom of plant overload or a common speed control problem on all engines. The
under frequency trip does very little for the redundancy concept because it is not fully
selective, this is usually a classification society requirement. In general, other protective
functions need to be present to prevent the power plant reaching this condition.
2.12.10 Loss of Excitation
This function is designed to prevent the generator running asynchronously and also to
prevent it becoming a significant VAr drain on the surviving generators which may cause them
to trip on over-current. This function may compound a failure in which an AVR on another
generator has failed to full excitation. In this case the healthy generators may be tripped by
their loss of excitation protection leaving only the faulty generator connected. The faulty
generator may then be tripped by its own protection resulting in blackout.
2.12.11 Reverse Power
This function is designed to prevent a faulty generator engine being motored by the other
online generators and becoming a significant kW load which may cause them to trip on over
current.
2.12.12 Phase Over Current
This function is designed to protect the generator from thermal damage associated with an
over-current condition. If the protective functions in the drives and power management
system are working correctly, healthy generators will not be overloaded. In systems designed
to operate at a high power factor, over-current and overload are almost the same condition.

62 IMCA M 206
2.12.13 Over Frequency
This fault could occur as the result of a severe engine governor failure which could cause the
entire bus frequency to rise to unacceptable levels. This function provides little benefit for
the redundancy concept other than to limit the potential for damage and ensure that
generators can be restarted after blackout.
2.12.14 High Set Over Current
This function is often used to trip the bustie breakers if a short circuit fault occurs on the
main switchboard when the vessel is operating with busties closed.
2.12.15 Trip Coil Monitor
This function will alarm if a faulty trip coil is detected on a generator circuit breaker. This is a
very useful function to prevent a hidden failure compounding another fault which could defeat
the redundancy concept.
2.12.16 VT Fuse Failure
This function alarms if a faulty fuse is detected on a generator VT. It provides further
confidence that the protection system is healthy. A faulty VT will also prevent a standby
generator synchronising and may cause certain meters and PMS indications to be in error.
2.12.17 Diode Failure
Loss of a single diode in an alternator does not usually cause the machine to fail immediately
but may reduce the effectiveness of the excitation system. In some protection schemes
detection of a faulty rotating diode initiates an alarm and a start request for a standby
generator to connect. In other schemes the faulty generator may be tripped.
2.12.18 Gen Winding RTD
High stator winding temperatures are indicative of a fault which may be related to overload
or a malfunction of the cooling system. In this protection scheme detection of high
temperatures initiates an alarm and a start request for all standby sets to be connected.
2.12.19 Gen Bearing RTD
High bearing temperatures are indicative of a fault in the bearing lubrication or that the
bearing itself is beginning to deteriorate. In this protection scheme detection of high bearing
temperatures initiates an alarm and a request for standby sets to connect. As a faulty bearing
can lead to seizure and loss of synchronism, the faulty machine should be taken out of service
as soon as possible.
2.12.20 Field Current
This is similar to the loss of excitation trip but for a less critical condition; this initiates an
alarm and a request for all standby sets to be connected.
2.12.21 Protection Relay Faulty
This alarm is initiated if the generator protection relay fails to pass its own internal diagnostic
check. It is a useful feature which improves confidence that the relay is healthy and thus
prevents hidden failures compounding a power system fault.
2.12.22 Lockout
This function acts to lockout generators from reconnection if they have tripped on a
particular fault. Generators should only be locked out from reconnection if they are actually

IMCA M 206 63
faulty. Care needs to be taken not to lockout generators for external faults, otherwise the
reliability and effectiveness of blackout recovery may be impaired.
2.12.23 Earth Fault Sometimes Called Zero Sequence
This function will alarm or trip the generator on detection of an internal earth fault. On high
resistance earthed power system, the earth fault current may be too low to activate the
differential protection; therefore a more sensitive form of protection is required. Restricted
or directional earth fault protection using core balance CTs may be used. A core balance CT
is a large current transformer which passes over all conductors in a three phase circuit. Care
should be taken when terminating the screen of a cable protected by a core balance CT; the
earthed screen should not pass through the CT.
2.12.24 Quick Trip
This function is a combination of:
high set over current;
negative sequence;
under/over voltage;
under/over frequency;
any other non-selective protective function.
It is intended to open the bustie to separate the two halves of the power plant before these
functions trip their respective generator circuit breakers. Although all generator protection
relays will see the fault, only the protection relays on the side with the fault will trip leaving
half the power system in operation. This function is only effective in splitting the plant, which
is effective for a redundancy concept with a 50% split. It is more difficult to use this function
effectively in a power system with a multi-way split.
2.13 Advanced Generator Protection (AGP)
2.13.1 General
Very significant advances have been made in the area of marine power plant protection in
recent years. Possibly the greatest improvement has come in the form of protection systems
able to identify which generator is responsible for causing a severe active or reactive power
sharing imbalance. Fuel control and excitation system faults are relatively common in marine
power systems. When governor and AVRs fail to an inert state (no fuel or no excitation) the
failure effects are relatively benign and will only result in the loss of the faulty generator,
provided the system is properly protected by traditional generator protection.
Unfortunately, traditional generator protection offers little if any protection against failures to
an active state (typically excess fuel or excitation), which can both result in severe load
sharing imbalance and blackout. In fact the incorrect response of traditional generator
protection is usually responsible for causing the blackout.
In many cases the driving force for developing this improvement has come from vessel
owners themselves, but protection systems are now available from several sources including
some of the large electrical system supplier and vessel automation vendors. All systems offer
certain core protection features but are also capable of being expanded to address many
other less critical failure modes and provide a useful backup to the traditional generator
protection.
2.13.2 Core Protection Functions
The two most important failure modes that must be protected against are failure to excess
fuel and failure to excess excitation. In the case of failure to full or excess fuel, there is a
risk in light load conditions that all healthy generators will trip on reverse power or over
frequency. In the case of exaction faults the healthy generators may trip on their field failure

64 IMCA M 206
protection or over voltage. In some cases the system may be left running on the faulty
generator which subsequently trips.
2.13.3 Predecessors to Modern AGP
A few vessels were fitted with bespoke protection systems of various degrees of
sophistication and early examples can be found dating back over 15 years.
Failure to excess fuel could be protected against by maintaining a base load in excess of the
rating of the largest generator. This was easily achieved by vessels using thruster bias mode
for DP control but was wasteful of fuel and not very environmentally friendly. In some cases
this protection measure was inadvertently defeated by the power management system which
was programmed to shed thruster bias and/or thruster load on detecting the overload of any
single generator.
Other early attempts to provide protection included systems which would trip the bustie if
two or more generators were simultaneously in the reverse power condition.
2.13.4 Principle of Operation
There are two basic principles of operation for advanced generator protection:
voting systems;
conformance to predicted generator behaviour.
Protection systems based on voting functions tend to be centralised protection systems
which collect and compare information on all online generators. They attempt to identify the
faulty generator by observing that one generator may acquire all the system active or reactive
power and other generators shed it. These systems assume the generator carrying all the
load is the faulty generator. Great care should be taken in the design of the plant to ensure
that this is the only failure mode that can have this effect otherwise the protection system
may trip the wrong generator. For example, if there are three generators online and two of
them shed load because they share a common faulty fuel system the generator left carrying
the system load is the healthy generator. If the protection system assumes that the generator
carrying all the load has taken it from the others it will trip the healthy generator leading to
blackout. The security of voting systems can be improved by combining the power
comparison function with bus frequency measurement. The principle of operation is that a
generator taking the system load from others because it has failed to excess fuel will drive up
the power system frequency. A generator which is taking load because others are shedding it
will not do this and bus frequency may fall.
Protection systems based on conformance to predicted generator behaviour do not require
information on the status or condition of other generators and can be designed such that
there is an independent protection system for each generator. The power plant is operated
in uncorrected speed and voltage droop mode so that the operating point of healthy
generators can be predicted from the speed and voltage droop characteristics. Figure 57
shows how the protection function works in the case of two generators running in parallel.
For as long as both generators are healthy, their operating points follow their respective
speed and voltage droop lines. When G2 fails to full fuel it takes load from G1 and drives it
into reverse power. However, because the speed (frequency) of both generators needs to
remain the same, the operating point of G2 deviates from the droop line. The operating
point of G1 on the other hand remains on the droop line. The protection function is
therefore created by placing a window round the droop line and tripping any generator that
strays outside it. Suitable delays are added to allow for system dynamics and the AGP must
act to isolate the faulty generator before the traditional generator protection operates to trip
the healthy ones.

IMCA M 206 65

Figure 57 Principle of AGP

66 IMCA M 206
3 Power Management
3.1 Requirement for a Power Management System
3.1.1 International Maritime Organization
IMO MSC 645 does not specifically require that a power management system is fitted, only
that it should be suitably reliable if it is.
Generally, power management systems should fail in such a way that:
the power plant continues to run as set;
local control is still possible.
3.1.2 Classification Societies
Most classification societies require that DP Class 2 and DP Class 3 vessels are fitted with a
power management system but the requirement to have such a system may be waived if the
various functions provided by a basic power management system are located in other
systems. Also, if there would be no significant advantage in having a power management
system, such as in the case of a DP vessel using direct diesel driven thrusters, it should also
be waived.
3.1.3 Power Management System Principles
The primary objective of a power management system is to ensure continuity of the power
supply to essential consumers under all defined operating conditions. In the event that it fails
in this objective it may be programmed to restore power automatically.
3.2 Power Management System Architecture
3.2.1 Centralised Control Systems
The older and more basic power management systems tend to be centralised control systems
in which a single processor or master slave arrangement interfaces to the entire power plant
from a single cabinet located in the engine control room. Such designs offer little in the way
of redundancy and the need to bring large amounts of analogue control signals back to a
single location can present difficulties for DP Class 3 designs. Variations on this design use
remote stations to gather the information from the field by way of a data communications
network, but all the control algorithms reside in the central controller.
3.2.2 Distributed Control Systems
Distributed control systems have been used in DP vessels for over a decade now and are the
preferred solution for larger more complex vessels. In distributed control systems the
control algorithms reside in processors located out in the field and close to the machinery
they control and monitor. Failures in these field stations generally affect only the machinery
controlled by that field station, but the effect of failures in field stations for power
management functions may be more widespread as the actions of one field station may
depend on receiving information from another.
3.2.3 Intelligent Consumers
There is an increasing trend to move the intelligence and control algorithms even closer to
the machinery they control. As many elements of the DP system such as engine and
thrusters are now controlled by local PLCs it is possible to put much of the intelligence
required to create an integrated propulsion system into the machinery itself. Features of
such intelligent consumers include the ability to:
make the equipment ready for remote control independent of any other control system;

IMCA M 206 67
take direct action to relieve stress in the power system;
monitor their own condition and disconnect themselves if they consider themselves to
be faulty.
3.3 Power Management System Hardware
3.3.1 PMS as Part of an Integrated Automation System
In large complex vessels, the power management system tends to be just another software
module within the overall integrated automation systems which may include:
dynamic positioning;
vessel management;
power management;
engine control and protection;
vessel safety system.
Figure 58 shows a few of the most significant interfaces between the IAS and the power plant.
MARINE
AUXILIARY
SERVICES
DP
INTEGRATED AUTOMATION SYSTEM
VESSEL
MANAGEMENT
ENGINE
CONTROL
AND SAFETY
POWER
MANAGEMENT
ENGINE
DRIVE
GOVERNOR
AVR
ENGINE
GOVERNOR
AVR
ENGINE
GOVERNOR
AVR
SYNC SYNC SYNC
SWITCHBOARD
DRIVE DRIVE
POWER
MANAGEMENT

Figure 58 Integrated automation system
3.3.2 PMS as a Standalone Function
In less complex vessels there may be no integrated automation system and it is not unusual to
find that the power management system is provided by a different vendor to the DP control
system and vessel alarm system. The power management system may be separate from the
control and alarm system or there may be a link to record PMS alarms. The PMS may only
offer very basic functions and load sharing. Engine control could be carried out by other
standalone hardware as shown in Figure 59.

68 IMCA M 206
MARINE
AUXILIARY
SERVICES
DP SYSTEM
CONTROL ALARM
MONITORING
ENGINE
CONTROL
AND SAFETY
ISOCHRONOUS
LOAD SHARING
ENGINE
DRIVE
GOVERNOR
AVR
ENGINE
GOVERNOR
AVR
ENGINE
GOVERNOR
AVR
SYNC SYNC SYNC
SWITCHBOARD
DRIVE DRIVE
POWER MANAGEMENT

Figure 59 Standalone PMS
3.4 Power Management Functions
3.4.1 Range of Available Functions
The extent of control provided by the PMS varies significantly between installations. The
types of functions considered to be power management functions also vary from system to
system and the boundary between PMS and vessel control functions may be determined by
the system architecture.
A large range of control options are available in the most sophisticated systems including:
engine control, monitoring and protection;
charge air temperature control;
HT cooling water control;
turbocharger assist control;
remote start and stop initiation of generator synchronising;
standby generator running order selection;
standby generator management standby time limitations;
load sharing (kW);
asymmetric or fixed target load sharing;
reactive power sharing (kVAr);
frequency control;
voltage control;
synchronising;
load dependant start and stop;
reactive power dependent start;
current dependent starting;
alarm initiated starting;
start all;
consumer load application rate control;
control of multiple independent power systems;

IMCA M 206 69
control of regenerated power;
consumer priority (drilling);
control of heavy consumer starting advanced power reservation;
blackout prevention phase back (load shedding) of heavy consumers;
load shedding of non essential consumers;
blackout restart of generators;
blackout recovery of power plant.
Very basic power management systems may offer little more than remote manual control of
generator starting, standby running order selection and load dependent starting. In such
installations it may be necessary to supplement the PMS functions with standalone protective
functions to ensure the power plant is fully protected against blackout.
3.4.2 Synchronising and Dead Bus Closing
This can be a PMS function or a switchboard function. Synchronisers are widely available
from a range of sources including engine governor manufacturers. These devices control the
engine speed to bring incoming alternators into phase, voltage and frequency alignment
(excitation control is not normally used in synchronisers for marine applications).
Classification societies usually require that these automatic synchronisers are backed up by a
manual circuit breaker control and a synchroscope or synchronising lamps. Generators are
normally connected to the bus while running very slightly above system frequency to ensure
that reverse power protection does not immediately trip the generator off line.
Synchronising may fail when attempting to connect generators during large load fluctuations.
Such conditions may arise when several generators trip because of a common mode failure
and blackout protection functions are actively shedding load to prevent remaining generators
from tripping on overload. Thus it is clear that synchronisers may be under the most
onerous duty just when standby generators are needed most.
Power management system interaction with the DP system could be used to freeze load for
long enough to allow successful synchronisation, however modern power system
manufacturers were confident of synchronising under all steady state and transient load
conditions. Generally, system manufacturers recommend one synchroniser per generator for
redundancy.
In modern power management systems it is possible to configure the function of a
synchroniser in software and use the PMS I/O modules to control the generators speed and
monitor generator and bus waveform alignment. The PMS sends the signal to close the
generator circuit breaker directly rather than just initiating the operation of the external
synchroniser.
Most synchronisers also have what is called a dead bus closing facility. This overrides the
synchronising function if the bus is dead. Power management systems may have an additional
feature designed to prevent two generators dead bus closing at the same time. This may be
achieved by staggering the start of generators following blackout recovery or using a system
which passes a token from one generator to the next over the vessel management system
network. Only the generator holding the token can dead bus close.
It is important that dead bus detection is robust and cannot incorrectly indicate that the bus
is dead when it is live. Dead bus connecting a generator when the bus is actually live can
cause very severe power system transients leading to blackout. Some power management
systems make use of several sources of information such as blackout relays and voltage
transducers, but in many cases these sources are connected to the same bus VT and interface
with the same I/O module and are not as independent as they appear; such an arrangement is
shown in Figure 60. Some power management systems also include the status of generator
and bustie circuit breakers in the detection algorithm to improve confidence in bus bar
status. Some designers use a second source of bus voltage data to confirm the bus VT signal.
Even where multiple sources are used there is a risk of spuriously declaring a blackout if

70 IMCA M 206
there is insufficient delay in the detection algorithm to cater for the effects of voltage dips
associated with short circuit faults.
BLACKOUT
WHEN K1-K3
CLOSED
VOLTAGE TRANSDUCER
BUS VT
K1
I/O MODULE
VOLTAGE
K2 K3
BUS
FREQUENCY VOLTAGE TRANSDUCER
K1
K2
K3

Figure 60 Blackout detection from single source
3.4.3 Frequency Control
In control schemes where the PMS acts to alter the governor speed set point as a means of
correcting droop, frequency control is the responsibility of the engine governor and the
power management system. Typically, the PMS interfaces to the governor by two digital
inputs, one for raise speed set point and the other for lower speed set point. This type of
interface was originally developed to allow power management systems and remote manual
controls to adjust the speed set points of mechanical governors by driving a speeder or pilot
motor clockwise or anti-clockwise. The speed set point will increment at a defined rate
during the time that the raise or lower contacts are closed. Thus, the speed set point is
effectively adjusted by the time for which the control pulse is applied. This type of interface
is still used on modern digital governors and is vulnerable to failures. Typical faults include
broken wires disabling the raise or lower signal. A more dangerous fault is when the
contacts for the set point raise signal stick in the closed position driving up the load on the
faulty generators and pushing the healthy generators towards the tripping point of their
reverse power protection.
When the PMS is responsible for frequency control and load sharing, computations for how
to adjust the governor to correct frequency and load sharing imbalance are carried out
simultaneously. A composite time pulse is calculated for each governor that will
simultaneously correct frequency and load sharing deviations.
3.4.4 Voltage Control
Voltage control by PMS is less common than frequency control. This is because several
power management system providers choose to leave this function under the control of
AVRs operating in uncorrected droop mode. Other power system manufacturers choose to
control AVR set points for voltage control and reactive power sharing in much the same way
as generator speed control. There is little evidence to suggest that compensating for voltage
droop offers any significant advantages for the redundancy concept. However, introducing
additional complexity into the generator control system can also introduce additional failure
modes.

IMCA M 206 71
3.4.5 Multiple Independent Power Systems
This refers to the need for a power management system to be able to manage bus sections as
separate power systems when opening and closing busties. Several classification societies
have requirements in their DP rules relating to the need for the PMS to be able to operate
effectively when the power system is subdivided into smaller independent power systems.
3.4.6 Load Dependent Start/Stop
This function is universal amongst power management systems. Although the need for load
dependent starting of standby generating sets is clear, vessel operators are divided as to the
wisdom of allowing generators to be automatically stopped. Many vessel operators prefer to
have the power management system alarm on light loading rather than act directly.
Classification society rules may require that there is a facility to disable the load dependent
stop function.
3.4.7 Blackout Restart
Blackout restart is an extremely important feature of any power management system but is
generally only present in the more sophisticated systems. There are many older vessels in
service where manual intervention is still required after a blackout. At least one classification
society makes reference to the requirement for connection of a standby set and the
sequential automatic reconnection of essential consumers within its DP rules.
A modern power management system should be able to re-establish limited propulsion
capability rapidly and automatically without any human intervention. As a first step, a typical
blackout recovery strategy will start all available generators. In some cases, the standby start
matrix may be disregarded and the first generator to run successfully will be closed on to the
dead bus bars. The actions that follow may be highly vessel specific but typically the power
management system will then close the HV and LV transformer feeder circuit breakers in
order to restore essential services. This action will be followed by a sequential restart of
services that were already running before blackout occurred. To prevent further problems,
the power management system will not attempt to reconnect any generator to a switchboard
that has tripped on a fault or start generators that were withheld from the standby start
matrix for any reason.
In some blackout recovery strategies, the first action of the PMS is to open all generator and
bustie circuit breakers. Thus it is important that the method for detecting blackout is secure
and reliable. In PMSs with this type of function, a false indication of blackout may actually
create a blackout.
There are conflicting opinions on the wisdom of automatically restarting thrusters and
assigning them to the DP system, due to the risk of repeating the fault that causes the power
system to blackout out in the first place. However, vessel owners who have such functions
have stated that the benefits of having the DP control system automatically arrest the drift-off
outweigh the possible risk that the protection system fails to isolate and lock out the faulty
consumer which caused the blackout.
3.4.8 Autonomous Restart of Thrusters
Possibly one of the greatest advances in recent years is in the ability of certain designs of
thruster drive to make themselves ready for DP independently of any centralised control
system. In the event of any significant disturbance or voltage dip which causes them to
disconnect, these thrusters will monitor the condition of the main power systems and
reconnect as soon as conditions permit. This makes blackout recovery of generators and
thrusters a parallel operation rather than a sequential one, saving a considerable amount of
time compared to conventional designs. Systems based on these principles can typically
recover from blackout in several tens of seconds rather than minutes. When attempting to
optimise blackout recovery times it is important to carefully consider the control sequences
leading to generator and thruster connection. In particular, it is important to identify any
unnecessary delays or permissive that might prevent a thruster connecting. For example, it
may be better to allow a thruster to start without its cooling fans running than lock it out

72 IMCA M 206
from reconnection until they become available, particularly if the thruster will take many
minutes to overheat without them, and there will be alarms to indicate the over temperature
condition.
3.4.9 Start Prevention of Heavy Consumers
This is not a standard feature on every power management system but is common on more
sophisticated systems. The additional load of, for example, a large deck crane may require a
standby generator to be connected if kW or kVAr spinning reserve will be compromised or if
the anticipated transient voltage dip threatens malfunction. A typical heavy consumer
function will allow the start to proceed after connecting the required number of generators
according to a look up table. As it is possible for multiple start requests to occur near
simultaneously, the heavy consumer module needs to be able to treat each request
separately.
3.4.10 Load Limitation/Load Shedding General
Load limitation/shedding may be carried out by several systems including:
the DP control system;
the power management system;
the generator protection;
the thruster pitch control system;
the thruster speed control system in the variable speed drive.
The DP control system is programmed to limit or reduce the thrust order if there is
insufficient power available. In the event of a shortfall in power, the DP control system will
generally sacrifice position to maintain heading. The load reduction functions in the DP
control systems are generally considered to be too slow to provide effective blackout
protection so it is normal to supplement these with fast acting load reduction functions in the
power management system. Such functions may also be included as part of the control
system for variable speed drives or for generator protection and these are discussed in more
detail in the sections that follow.
Care needs to be exercised in the design and tuning of load limitation and load shedding
schemes as poorly designed systems can induce severe oscillations in the power systems
which can escalate to blackout or prevent standby generators connecting.
Careful co-ordination is required to ensure that load reduction functions and load dependent
start function are properly co-ordinated. Load dependent start should be initiated at a power
level below the point at which load reduction or load shedding measures are activated.
Failure to arrange the levels correctly creates a condition where standby generators cannot
start because the load is being maintained below the starting point. Such a condition will
result in a drift off unless the engineers connect additional generators manually.
3.4.11 Load Shedding Power or Frequency
Load reduction systems may be based on:
available power;
power system frequency.
Available power: Load reduction systems based on available power monitor the power
being supplied by the generators and status of generators (online or offline) to determine the
level of available power. Such systems depend upon the generators being able to deliver
rated power and are thus vulnerable to failures caused by conditions such as fuel starvation
where the generators cannot deliver rated power. Care is required to ensure the power
available calculation is robust and some PMS manufacturers carry out confidence checks by
comparing the power delivered by the generators to the load consumed load reduction
functions are suspended if a significant discrepancy is detected. On older vessels the power
limitation signals connect the PMS field stations to the thruster or heavy consumer control

IMCA M 206 73
system by way of dedicated analogue links. This arrangement ensures the necessary speed of
load reduction but may introduce complications in DP Class 3 designs where the effects of
fire on cable route must be considered. In more modern vessels the load reduction signals
are carried by the dual redundant communications network.
Power system frequency: Load reduction systems based on bus frequency simply reduce
load to prevent unacceptable drops in bus frequency which indicate that the generators are
unable to support the load. Such systems are not vulnerable to conditions which limit the
power of the generators but care must be taken to ensure the frequency measuring scheme
is robust and not likely to trigger a spurious load reduction due to erroneous frequency
measurements. Detection of low frequency can also be used to open busties and trip non
essential consumers.
3.4.12 Load Shedding Systems in Variable Speed Drives
The control system for variable speed drives used for thrusters, cranes and pipelay systems
may also contain independent load reduction functions based on measuring the frequency at
their power input. Because each drive makes its own frequency measurements, failures are
generally limited to one consumer. As this makes for a very robust system some vessel
owners choose to base the entire load reduction system on this principle and have no other
system based on centralised control and measurement other than the DP control system.
Careful design and tuning of these functions is important to prevent unwanted oscillations.
Some systems use a fast-attack, slow-release function. In this type of function a very rapid
power chop is initiated followed by a slow ramp up to the limit of the available power
determined by the bus frequency.
3.4.13 Load Limitation/Shedding of Variable Speed Drives Based on Available Power
The speed with which variable speed drives can be controlled allows power management
systems to use them for blackout prevention purposes. This form of control is often based
on measurement of consumed power and estimates of available power based on generator
rating.
The control lines for load limitation are usually a mixture of 4-20mA loops and net I/O. In
some cases analogue load limitation signals will connect PMS field stations directly to the
variable speed drive control systems. In other cases the PMS field stations will communicate
with the thruster or drive field station which will then pass the signal or (some derived signal)
to the variable speed drive itself.
In properly designed systems the effect of PMS failure or control wiring failure on thrusters
will be fails to no limit with appropriate alarm. The failure modes for non DP consumers
may be to full limit or no limit, depending on what is considered to be the safest mode.
Figure 61 shows the signals related to power limitation and control.
VFD
SPEED OR TORQUE CMD
POWER AVAILABLE
(OR LIMIT)
POWER CONSUMED
POWER LIMIT ACTIVE

Figure 61 Signals relating to load reduction

74 IMCA M 206
3.4.14 Load Limitation and Reduction Functions
To ensure that available power is shared in an orderly manner among all the various
consumers according to their importance, each consumer is sent a power available signal
that tells the drive how much power it can have at any instant in time. If the speed or torque
command to that drive requires more power than the power available signal allows, the drive
will set its power limit active signal to advise the PMS and DP system that it has reached the
power limit. The power available signal is recalculated each scan cycle of the PMS and can be
visualised as a dynamic ceiling on the power that the drive can draw. When there is plenty of
spinning reserve, the power limit will be high and well above what the drive is actually
consuming, but as the spinning reserve is used up the ceiling will fall to just above what the
drive is actually drawing. The power available signal can be absolute, i.e. the total power
that the drive can draw, or relative, the amount by which it can increase its load.
The example below shows a relative figure for power available.
If the demand for power exceeds the power available signal, the effect is that the drive load
ramps up by the value of the power available signal in each scan cycle until the desired load
level is achieved or bus capacity limits further increase. In real systems it is possible to
specify an additional ramp function to prevent significant dips in bus frequency and voltage or
to prevent excessive torque in the drive train. It is also possible to specify a delay on
reallocation of power after a power reduction is active (such features are not included in the
discussion below).
To prevent an overload created by several systems, each simultaneously grabbing all the
spinning reserve, the power available calculation for each consumer includes coefficients
reflecting the relative priority of consumers.
Two conditions are simultaneously monitored and controlled:
consumer wishes to draw more power than has been allocated to it;
overload caused by sudden loss of generating capacity.
3.4.15 Power Available for Load Increase Example System
The example Figure 62 is in a much simplified power system with two generators, one
thruster, a crane and a fixed hotel load. There is no control over the hotel load but both the
thruster and the crane can be controlled as described above.
G1
5 MW
G2
5 MW
HOTEL
CRANE
4 MW
3 MW 1 MW

THRUSTER
1 MW

Figure 62 Example power system

IMCA M 206 75
The crane and the thruster are each allocated power according to the following functions:
power available to crane = crane allocation x (bus capacity x crane priority bus load);
power available to thruster = thruster allocation x (bus capacity x thruster priority bus
load).
The variables and coefficients are defined as follows:
bus capacity is equal to the sum of the ratings of the online generators;
bus load is the total load on the bus;
crane priority is 90% the crane cannot raise the bus load to more than 90% capacity if
the bus load exceeds 90% the crane load will be reduced;
thruster priority is 100% the thruster can raise the bus load to 100% capacity and will
force down the crane load to ensure it can access the power it needs to hold station;
crane allocation is 40% at any instant in time, the crane can increase its load by 40% of
power available;
thruster allocation is 60% at any instant in time, the thruster can increase its load by 60%
of power available.
Note that in commercially available power reduction schemes the control functions are more
sophisticated with additional coefficients. This much simplified example is only intended to
help the reader understand the concept.
3.4.16 Thruster Priority
Because the thruster has a higher consumer priority than the crane it will continue to be
assigned a power available figure which allows it to increase bus load over the crane priority.
This has the effect of invoking the crane overload function which then ramps down the crane
load incrementally as thruster load increases.
3.4.17 Consumer Allocation
The effect of the consumer allocation is that whatever power is available for simultaneous
load increase by thruster and crane, the crane can have 40% of it and the thruster can have
60%. The consumer priority and allocation can be set to suit a particular application.
3.4.18 Response to Overload
In addition to managing the rate and percentage by which each consumer can increase its
consumption, the PMS must respond rapidly to loss of generating capacity before the
surviving generators trip on overload. Various schemes are possible but load shedding in
response to bus overload and falling bus frequency are almost always included. Load
shedding in response to individual generator overload may also be included but this needs
careful scrutiny to make sure it will not compound a governor failure to excess fuel leading to
blackout.
Load shedding in response to bus overload is performed using the following functions:
overload thrusters = bus load thruster priority x bus capacity;
overload crane = bus load crane priority x bus capacity.
3.4.19 Actual Vessel Systems
Reference to philosophy documents for actual vessel systems will reveal more complex
algorithms which allow the user to assign the percentage of available power that is allocated
from each bus in a multi-bus system and also to include a fixed power reserve for such things
as active heave compensation, but it may be useful to consider the simplified scheme in Figure
62 prior to considering the effects of additional complexity.

76 IMCA M 206
3.4.20 Example Power Limitation and Reduction Scheme
Refer to Figure 63. In the example power system below, the generators are rated at 5MW
each and the figures below each consumer are the power consumption used in the first
example rather than their rating. Two examples are used to demonstrate how the load
limitation and reduction systems respond to changes in the power system conditions to
prevent blackout and drift off. The figures used are intended to help the reader understand
the action of the power management system rather than represent realistic conditions.
3.4.21 Example 1 Sudden Loss of Generating Capacity
In the first example, the hotel load is 1MW and the thrusters are consuming about 3MW but
this is varying up and down by about 1MW due to the environmental conditions, as shown in
Figure 63 such that the total bus load is peaking at 9MW or 90% of generating capacity. The
crane is hoisting at a rate that requires 4MW.
At 45s in Figure 64, one of the two 5MW generators trips on a fault. Immediately the
thruster and crane loads are each reduced by the calculated overload, but are allowed to
ramp up again. However, as the new generating capacity is not enough to support the
original thruster and crane loads, the crane is forced to slow down during the thruster power
peaks. As Figure 65 shows, the crane and thruster power available (PA) figures fall almost to
zero and the bus load is near to the remaining generating capacity of 5MW.
0 50 100 150
0
2
4
6
8
10
x 10
6
Thruster Load, Crane Load, Hotel Load
Time(s)
P
o
w
e
r

(
W
)
Total Bus Load
Thruster Load
Crane Load
Hotel Load

Figure 63 Initial thruster, crane and hotel loads


IMCA M 206 77
0 50 100 150
0
2
4
6
8
10
x 10
6
Thruster Load, Crane Load, Hotel Load
Time(s)
P
o
w
e
r

(
W
)
Thruster Load
Crane Load
Bus Load

Figure 64 Thruster, crane and hotel loads after loss of one 5MW generator
0 50 100 150
0
2
4
6
8
10
x 10
6
Bus Load, PA - Crane, PA - Thrusters
Time(s)
P
o
w
e
r

(
W
)
Bus Load
PA Crane
PA Thrusters

Figure 65 Effect on power available signals to thruster and crane
3.4.22 Example 2 Sudden Demand for Thrust
In the second example, bus capacity is 10MW and the thrusters are idling at a steady
1.5MW, with the crane hoisting at rate which requires 4MW. The hotel load is 1MW.
At 90s into Figure 66 there is a sudden demand for the thrusters to increase to 8MW.
The power limitation functions limit the rate at which the thrusters can ramp up and the
crane load is shed away such that the crane has to slow to a stop. Figure 67 shows how
the bus load levels out at 9MW and the remaining power available to the thrusters is
low. Although there is some bus capacity left this is not available to the crane as its
priority level is 90%.

78 IMCA M 206
Had the thruster demand been 5MW instead of 8MW, as shown in Figure 68, then 3MW
would have been available to allow the crane to continue to hoist.

0 50 100 150
0
2
4
6
8
10
x 10
6
Thruster Load, Crane Load, Hotel Load
Time(s)
P
o
w
e
r

(
W
)
Thruster Load
Crane Load
Hotel Load

Figure 66 Sudden demand for thrust
0 50 100 150
0
2
4
6
8
10
x 10
6
Bus Load, PA - Crane, PA - Thrusters
Time(s)
P
o
w
e
r

(
W
)
Bus Load
PA Crane
PA Thrusters

Figure 67 Bus load and power available following large demand for thrust

IMCA M 206 79
0 50 100 150
0
2
4
6
8
10
x 10
6
Thruster Load, Crane Load, Hotel Load
Time(s)
P
o
w
e
r

(
W
)
Bus Load
Thruster Load
Crane Load
Hotel Load

Figure 68 Thruster step load of 5MW
3.4.23 Load Shedding Preferential Trips
This feature is universal within all the more sophisticated power management systems but
may be provided by switchboard protection functions in some cases. The effectiveness of
preferential tripping in some DP vessel types may be limited by the size of non-essential loads
that can be made available for tripping.
Typical non essential loads include:
ventilation fans;
HVAC;
galley services;
calorifiers;
water makers;
compressors.
3.4.24 Drilling Priority
In the case of vessels operating in deep water where the permissible position excursion is
relatively large, some vessel owners have chosen to prioritise power to drilling and pipelay
systems until a defined excursion is reached. At this point, priority for power is returned to
the thrusters and the drift off will be halted. This small excursion is tolerated during the time
that the power shortfall is made good by the connection of standby sets. This arrangement
requires the DP control system to supply position information to the PMS so that it can
reconfigure.
3.4.25 Active Heave Compensation Reserve
More sophisticated power management systems offer a facility to cope with consumers that
can vary their power demand over a very large range in a short period of time. Active heave
compensation systems for drilling are typical of this type of consumer. In simple terms, a
false load figure equal to the expected peak power demand is added to the actual load to
maintain a power reserve and trigger load dependent starting to ensure that power reserve is
maintained. Thrusters and other consumers cannot access this reserve until they are given
power priority over the active heave compensation.

80 IMCA M 206
3.4.26 Harmonic Filter Control
Systems that employ harmonic filters can have these devices connected or disconnected by
the power management system as conditions dictate. Filter disconnection may be desirable
to improve power factor under certain running conditions or reduce power losses in
conditions where they are not required. A complementary strategy offered by one
manufacturer starts an additional generator when total harmonic distortion exceeds a certain
level.
3.4.27 Derating
Derating is a useful but unusual power management system feature when a diesel engine is no
longer able to attain its rated power. This feature allows the operator to reassign a new
maximum kW rating to one generator. The power management system will then use this
new value to determine the division of load between generators.
3.4.28 Starting of Standby Generators on Alarms from Running Generator
Connecting a replacement generator on detection of an alarm condition on a running
generator is another standard feature of most PM systems. Users may have the choice of
programming the automatic stopping of faulty sets depending upon the severity of the alarm.
Some vessel owners prefer to have manual stopping of the faulty generator.
3.4.29 Remote Start Stop
Almost all power management systems offer this feature, how it is actually achieved may
depend on the particular system configuration. For example, use of an engine manufacturers
control system or direct control from the power management systems.
3.4.30 Standby Generator Selection
The standby starting matrix is another central component in most power management
systems. This may be configured by the engineer to take account of ongoing maintenance.
Several incidents have been recorded where failure of a standby generator start was due to
improper setting of the starting order. Power management system manufacturers would
claim that these problems have now been overcome. Even in the most modern systems it
may be possible to select the same standby order number for more than one generator
without any alarm on the power management system. The PMS will generally connect the
generator with the lowest physical number. However this generator may be on a different
switchboard to the one the operator intended to connect.
3.4.31 Circuit Breaker Open/Close
Circuit breaker open/close is a standard feature of many systems and a network mimic is
often provided as an aid to correct selection. Interlocking of circuit breakers is normally
hardwired. In some cases software interlocks are used as a backup to the hardwired
interlocks. Some power management systems offer a convenient feature that allows the
power system to be configured for different types of operations.
3.4.32 Password Protection
Password protection on some systems controls access to more advanced features. When a
choice of settings is provided, some manufacturers will ensure that it is not possible to
inadvertently set parameters beyond reasonable limits.
3.4.33 Power Supply Arrangements for PMS
IMCA M 196 Guidance on the design, selection, installation and use of uninterruptible power
supplies onboard vessels provides details of UPS power supply arrangements for DP vessels
and further discussion of battery systems is provided in Section 4.5 of this document.

IMCA M 206 81
In older DP vessel designs it was not uncommon to find a single power management system
controlling the whole plant powered from a single UPS. This design was justified on the
grounds that total failure of the power management system left the power plant operating in
a stable condition and local manual control was still possible. In modern designs the power
management system is generally split along the lines of the overall split in the redundancy
concept and each part of the power management system will have its own independent UPS
supply.
Where the PMS is part of a large integrated vessel management system the PMS may share
UPS supplies with the rest of the vessel management systems. However, the number of UPSs
and their power supply source should be arranged to align with the overall split in the
redundancy concept.

82 IMCA M 206
4 Power Distribution
4.1 Power Distribution Schemes
4.1.1 General Principles and Influence on Redundancy Concept
The power distribution scheme of a DP Class 2 or DP Class 3 vessel forms the very heart of
the vessels DP redundancy concept. How generators, thrusters and auxiliary services are
divided into groups largely determines the vessels worst case failure and therefore its post-
failure DP capability. It is common practice to describe the electrical part of the redundancy
concept in the form of an overall power system single line diagram.
4.1.2 Design Methodology
There are several possible approaches to DP vessel design but two possibilities are outlined
below:
If the intention is to build a vessel similar to a unit that already exists in the fleet, then
data on the station keeping capability of this vessel may serve as a guide or baseline for
development of the new vessel;
If the new vessel is significantly different, then the process leading to development of the
redundancy concept follows the pattern below.
The hull form required for the application is chosen, typically ship-form, barge or semi
submersible. Catamaran forms are now becoming popular for certain applications.
Established regional data on weather, sea state and tidal currents is consulted to further
refine the choice of hull design and propulsion machinery.
For DP Class 2 and DP Class 3 vessels, the post worst case failure capability defines the
environmental conditions in which the vessel can operate and this can be related to uptime
by reference to the regional environmental data.
How the post worst case failure DP capability is achieved is determined by the redundancy
concept and worst case failure design intent which are intimately linked to the arrangement
of main and auxiliary switchboards. This will influence the type and rating of propulsion
machinery along with estimates of power requirements for hotel and process loads such as
drilling, pipelay, ROV support, etc. An allowance for maintenance of major machinery may
also be included.
The whole process may go through several iterations leading to the point where the DP class,
redundancy concept, operating configuration, WCFDI, single line power system diagram and
the rating and type of the major propulsion machinery has been established.
4.1.3 Distribution System Components
The power distribution scheme consists of:
switchboards with switchgear, protection and control equipment;
service transformers;
distribution boards;
motor control centres;
cables cable trays.
The purpose of the power distribution scheme is to distribute power to consumers at the
required voltage level.

IMCA M 206 83
4.1.4 Voltage Distribution Levels
There may be several voltage levels in a power distribution scheme to provide power at a
voltage appropriate to the type and power rating of the consumer. As might be expected,
low power consumers tend to be supplied at lower voltage levels than high power
consumers. As power ratings increase, the rated voltage is increased to reduce the current
and thus the cross section of the cables and windings required to supply the power.
Increasing the voltage rating also reduces the maximum fault current. Low voltage supplies
may be provided for safety reasons and because there are fewer restrictions on the
qualifications of maintenance personnel. In a large diesel electric vessel the distribution levels
might be as listed in Table 7. Voltage distribution levels for smaller or less complex vessels
are given in Table 8. However, some relatively large vessels now have LV power generation.
Voltage level designation Typical voltage Use
HV 11kV 6.6kV
Generators, thrusters, service
transformers, drilling and pipelay
drives
LV 690V 480V 440V 380V
Pumps, fans, service transformers,
HVAC, accommodation heating.
Emergency services
Lighting and small power 220V 208V 110V Lighting, single phase motors
Battery systems 110Vdc 24Vdc
Systems for navigation, DP vessel
management, thruster drives, engines
and switchboards
Table 7 Voltage distribution levels (large vessels)

Voltage level designation Typical voltage Use
Main generation level 690V Generators, thrusters, large pumps
Auxiliary system level 480V 440V 380V
Pumps, fans, service transformers,
HVAC, accommodation heating.
Emergency services
Lighting and small power 220V 208V 110V Lighting, single phase motors
Battery systems 110Vdc 24Vdc
Systems for navigation, DP vessel
management, thruster drives, engines
and switchboards
Table 8 Voltage distribution levels (smaller vessels)
4.1.5 Choice of HV or LV Power Systems
The choice of whether to specify an HV or LV power system depends to a large extent on
the installed power. Fault current levels may make an LV installation uneconomic or
impractical. For vessels with installed power above a figure of around 10MW an HV
installation is generally more cost effective. Below 10MW an LV installation may be an
attractive solution. The relatively higher costs of HV equipment may be offset by smaller
cable cross sections and switchboards with lower fault withstand ratings. Care must be taken
when specifying an LV system to ensure that there are no unacceptable restrictions on plant
configuration. For example, it may not be possible to connect all generators in parallel in
some LV designs as the fault rating of the switchboard would be exceeded. The need for
specialist maintenance engineers is sometimes cited as a reason for specifying LV installations.
4.1.6 Diversity
Diversity factor is the term used to indicate what fraction of the total connected load can be
simultaneously supplied by a switchboard or distribution board. Generally, main
switchboards must be rated for the total current that can flow through them. However, for
other parts of the distribution system, it may be accepted that not all the connected loads
will be operating at the same time and the distribution system may be designed to supply 75%

84 IMCA M 206
of the connected loads simultaneously, for example. Some designers choose to use the vessel
management system to monitor the load on the distribution system. A request to connect a
load results in an alarm indicating that there is insufficient capacity to complete the
connection without tripping the switchboards over current protection. The engineer can
then choose to stop some non-essential equipment to allow the connection to proceed.
Other designers choose to link the overload protection to the non-essential trips so that
load is automatically shed to prevent an overload.
4.1.7 Emergency Generators
Vessels which fall under the SOLAS regulations must have an emergency source of power
capable of supplying designated emergency consumers. In larger vessels this requirement is
typically fulfilled by the provision of an emergency generator. This generator and its
associated emergency switchboard form part of the vessels low voltage scheme but there are
requirements for separation between the two systems. On smaller vessels the emergency
generator may be 300kW but on larger vessels it is common to find generators of 1MW to
2MW rating.
It is sometimes possible for DP vessels to obtain an exemption from having an emergency
generator if they meet certain requirements in addition to their power system redundancy.
However, the emergency generator has so many useful functions related to black starting the
vessel, providing an alternative source of power and providing long-term backup to battery
systems that most DP vessel power plant designers choose to incorporate an emergency
generator in their design. Some even elect to have a full sized main generator as the
emergency set.
SOLAS and classification society rules influence the way in which the emergency generator is
used for non-emergency functions. Generally, it is accepted that the emergency generator
can act as a harbour set for short periods of time. This is useful if the harbour load is very
small in relation to the load at sea as it may help to prevent a large main engine running at
low load for an extended period.
Some classification societies interpret SOLAS requirements as:
Normal operation of the vessel shall be possible with the complete emergency electrical
power supply system out of operation;
All consumers that support functions required to be available in normal operation, shall
be supplied from distribution systems independent of the emergency electrical power
supply system;
All consumers required to be available in emergency operation shall be supplied from
distribution systems independent of the main electric power supply system;
Consumers required to have both main and emergency supply shall be supplied as
required by relevant rules applicable for these consumers. The primary supply shall be
from the main system. Upon failure of any of the required power supplies, an alarm shall
be initiated.
This last clause is generally interpreted to mean that mains power for UPSs and battery
systems required to support DP essential consumers should be supplied from the same side
of the main power distribution system as the equipment they support but can have a backup
supply from the emergency power distribution. Under no circumstances should all DP UPSs
and battery systems have a sole source of power from the emergency switchboard although
this arrangement was once popular.
In some cases class may object to the automatic transfer of a large non-emergency load to
the emergency switchboard if there is a possibility of overloading it.
4.1.8 Shore Power
Most designers choose to incorporate a shore power connection for convenience during dry
docking. This feature allows the LV power system to be run from the dock supply to provide
hotel services such as lighting, power sockets and ventilation. Shore power connections may
have interlocks and intertrips to prevent inadvertent paralleling of the vessels power plant

IMCA M 206 85
with shore power (see 4.1.12). Occasionally, specifications for MODUs include a
requirement for permanent shore based power as used by some fixed installations in the
North Sea. This is generally considered to be impractical.
4.1.9 Service Transformers
Service transformers are provided to transform power from one voltage level to another.
They are typically a three phase two winding transformer with a delta/delta or delta/star
arrangement with or without an earthed star point, as shown in Figure 69. Occasionally a
three phase transformer is constructed from three single phase transformers and a fourth
single phase transformer is provided as a spare winding.
R Y B
R Y B R Y B
DELTA / DELTA DELTA / STAR
R Y B

Figure 69 Service transformers
The voltage at a transformer secondary winding can be predicted from a knowledge of its
primary voltage V
1
and turns ratio N
1
/N
2
according to Equation 8.
1
2
2
1
2
1
I
I
N
N
V
V

Equation 8 Transformer formula
Service transformers can be water cooled or air cooled. Air cooled transformers can be
convection (naturally) cooled or forced cooled using fans. Some transformers may be
arranged to trip on over temperature and others may only have an alarm. It is not
uncommon in older DP Class 2 designs to find all the vessels service transformers in a single
compartment. This is acceptable but the provision of ventilation supplies and fire dampers
may need careful consideration.
Some classification societies may require over voltage protection on the transformer
secondary in view of the risk of an over voltage being coupled through from the primary side
although an earthed screen between primary and secondary windings may be accepted in
mitigation. Where an earthed neutral is provided on the secondary side, it may be necessary
to have only one service transformer earthed at any one time to prevent circulating currents.
Service transformers can have an unexpected influence on the redundancy concept of a DP
vessel particularly when they have cooling or other auxiliary functions which are powered
from sources other than the distribution systems they feed. Such arrangements are
surprisingly common and can cause distribution system failure effects to be more severe than
would otherwise be expected if not identified in the DP system FMEA.

86 IMCA M 206
Where the secondary side of two service transformers can be connected by a tie line
connecting their respective LV switchboards, both transformers will be able to contribute
current to any fault. In some distribution systems the LV switchboards are not rated for this
level of fault current and interlocks are provided to prevent the parallel operation of the
service transformers. Short-term paralleling may be arranged to allow bumpless transfer of
the load from one service transformer to the other.
Large service transformers may have a significant inrush current at the point of connection
which is capable of tripping over current protection. This can have a distinct disadvantage for
blackout recovery as it may be necessary to wait for two generators to connect before the
service transformers can be energised.
The inrush current transient is cause by the action of synchronising the magnetic core of the
transformer with the applied voltage at the time of connection which may drive the core into
saturation initially.
This effect can be prevented by using a pre-magnetising transformer. This is a small LV
transformer which is used to back feed the large service transformer from its secondary side
before the primary side of the service transformer is connected. Thus the magnetic core is
already in synchronism with the primary side supply when the transformer is connected.
4.1.10 Phase Shifting Transformers
Phase shifting transformers are normally used in marine power plant applications for
harmonic cancellation. Excessive levels of low order harmonic distortion are associated with
equipment malfunction, unwanted heating, audible noise, lighting failures and other faults.
The idea that a distorted waveform can be represented by a series of sinusoidal waveforms is
credited to Joseph Fourier, an 18
th
century French mathematician. The frequency of each
harmonic component is a multiple of the fundamental frequency which is typically 60Hz in
marine power systems.
Classification societies set various limits for Total Harmonic Distortion (THD) but figures of
5% for THD and 3% for the maximum contribution of any one harmonic are not uncommon.
In marine applications the primary sources of harmonics are from AC and DC variable speed
drives used for thrusters and drilling equipment. Figure 70 shows the distortion in the red-
yellow line voltage of a large DP vessel using twelve-pulse thruster drives.
Variable speed drives using three phase bridge rectifier front ends are still popular for
thruster drives and drilling equipment. These drives draw non-linear current with significant
quantities of 5th and 7th order harmonics. The level of low order harmonics can be reduced
by using two six-pulse rectifiers supplied from a single three winding transformer with a delta
primary and two secondary windings. One secondary is a delta and the other a star. This
arrangement introduces a 30 phase shift between the current pulses drawn by the two six-
pulse rectifiers and the dual rectifier drive is referred to as a twelve-pulse drive. Figure 71
shows the harmonic analysis of the waveform in Figure 70. Although it still contains
significant quantities of 11th and 13th harmonics, the 5th and 7th order harmonics have
almost been eliminated.
The increasing use of drives with active front end (AFE) rectifiers has removed the need for
phase shifting transformers but an isolation transformer is still required.

IMCA M 206 87
Timed event at 14/02/01 12:57:36.000
CHA Vol t s
12:57:36.000 12:57:36.005 12:57:36.010 12:57:36.015 12:57:36.020
Volt s
- 10000
- 7500
- 5000
- 2500
0
2500
5000
7500
10000

Figure 70 Harmonics distortion from twelve-pulse drives on 6.6kV marine power systems
Total RMS: 6609.47 Volts
DC Level : 4.64 Volts
Fundamental(H1) RMS: 6602.38 Volts
Total Harmonic Distortion (H02-H50): 4.48 % of FND
Even contribution (H02-H50): 0.18 % of FND
Odd contribution (H03-H49): 4.48 % of FND
Timed event at 14/02/01 12:57:36.000
CHA Vol t s
Thd H10 H20 H30 H40 H50
%of FND
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5

Figure 71 Harmonic content of 6.6kV voltage waveform
It is possible to create transformers with other phase shifts to further reduce the harmonic
contribution from drives. The International Electrotechnical Commission (IEC) devised the
concept of the transformer vector group code to indicate the internal connections of
transformer windings. The vector group takes the form of a three element code such as
Dy11.
D indicates the HV winding is a delta;
Y indicated the LV winding is a wye (star);
11 indicates the voltage of the LV winding leads the HV winding by 30 (by reference to a
clock face) in a Dy1 transformer the phase shift would be 30 lag.

88 IMCA M 206
Note the IEC standard refers to HV and LV windings not primary and secondary windings.
Transformers built to ANSI standards may have a vector diagram rather than a vector group
code on their nameplate.
4.1.11 Protection for LV Distribution
Protection system for LV distribution (where it is not the main power generation level) tend
to be limited to:
over current (short circuit);
overload for motors;
under voltage trip for service transformer feeders with suitable delay;
earth fault alarm.
At the lowest power distribution levels such as 120V or 24Vdc it is not uncommon to find
that distribution boards are populated with miniature circuit breakers. The selectivity
provided by these devices needs very careful scrutiny, particularly if the redundancy concept
depends on their correct operation.
4.1.12 Interlocks
Interlocks in power distribution systems are necessary to prevent the plant entering a
dangerous condition or configuration. Hardwired interlocks are required to prevent such
things as:
switchboards being configured in such a way that their short circuit rating is exceeded;
bus sections being connected when not in synchronism;
generators being connected when not in synchronism;
circuit breakers being withdrawn or engaged when closed;
generators being started or exciting when earth switches are engaged;
to prevent a variable speed drive being connected without pre-charging its DC link;
to prevent all generators connecting to a switchboard at the same time if these would
exceed the short circuit rating.
Interlocks are also used to restrict access to live parts when maintenance is being carried out.
Specialised key type interlocks are used to ensure that:
the switchboard earth switch cannot be engaged unless all generator and bustie circuit
breakers are in the open position;
cabinet doors for dual fed consumers cannot be opened unless both sources of power
are isolated.
Key type interlocks may have a mechanical part and an electrical part. The mechanical part
may be intended to physically lock the circuit breaker in the open position and the electrical
part will interrupt the closing circuit and apply a trip signal to open the circuit breaker if for
any reason the circuit breaker is closed.
Some interlocking schemes include inter trips. An inter trip is a signal sent to one circuit
breaker from another. Typical uses of inter trips are:
to open the HV bustie if the LV bustie trip, in power system that run with service
transformer in parallel, as shown in Figure 72;
to open the primary side circuit breaker of a service transformer if the secondary side
trips on a fault, as shown in Figure 73;
to prevent service transformers being operated in parallel, as shown in Figure 73;
to prevent the shore supply being paralleled with the vessels generators or to
disconnect all the service transformers from the LV power system if the shore supply is
connected as shown in Figure 74.

IMCA M 206 89
Care needs to be taken in the design of interlocks and inter trips not to create single point
failures that can cause a blackout. As the electrical parts of interlocks and inter trips often
form a common point between redundant parts of the power system there may be a
possibility that faults on the interlocking lines affect both power systems. In particular, the
effects of fire and flooding on interlocking lines need to be considered.
G G G
HV
G
LV
TRIPS HV BUSTIE
IF LV BUSTIE OPENS.
PREVENTS LV BUSTIE CLOSING
IF HV BUSTIE IS OPEN.

Figure 72 Inter-trip for parallel service transformers
G G G
HV
G
LV
TWO OUT OF THREE INTERLOCK
PREVENTS BOTH SERVICE
TRANSFORMERS OPERATING IN
PARALLEL
INTERTRIP INTERTRIP

Figure 73 Interlocks to prevent parallel service transformer operation
G G G
HV
G
LV
SHORE
SUPPLY

Figure 74 Shore supply interlock

90 IMCA M 206
4.2 Power Plant Configurations
4.2.1 Power Plant Topologies
There are a huge number of variations in the design of diesel electric power plant for DP
vessels although many designs are variations on a smaller number of basic designs and some
use elements from several different design concepts. All the examples which follow can be
built as DP Class 2 or DP Class 3 designs and can be operated as common or independent
power systems with suitable protection schemes.
G G
T3
T4
T5
T6
T2
T1
G
G G G

Figure 75 Asymmetric two-way split
Figure 75 shows one of the most basic arrangements. Each of the two main switchboards
connects to the same number of generators and thrusters but because of the arrangement of
thrusters in the hull, each main switchboard supplies different numbers of bow and stern
thrusters. Although very popular, this design does not make optimum use of the available
machinery because the worst case failure effect is loss of 50% of the generating capacity.
The worst case failure also leaves only one bow thruster operating which results in a poor
post-failure DP capability.
Figure 76 shows another popular arrangement. This design is based on a three-way split.
Provided the power system is properly designed and protected, the worst case failure is now
reduced to 33% of the power generating capacity and there should always be two bow
thrusters available after any single failure. However, this arrangement can also be operated as
two independent power systems in an asymmetric two-way split with only two generators
online. In this configuration four thrusters can be lost if one generators trips. This is a poor
operating configuration for DP Class 3 and designers wishing to adopt the three-way split
should consider the need for low load running to avoid the asymmetric configuration.
G G
T3
T4
T5
T6
T2
T1
G G
G G

Figure 76 Three-way split

IMCA M 206 91
G G
T3
T4
T5
T6
T2
T1 G
G
G G

Figure 77 Transferable or dual fed thrusters
Figure 77 shows another alternative. In this arrangement the power generation and
distribution system is arranged as a two-way split but two of the thrusters can be powered
from either main switchboard. The worst case failure is loss of 50% of the power generating
capacity but it should be possible to have two bow thrusters available after any single failure.
It is possible to provide every thruster with a dual feed such that the worst case loss of
thrusters is reduced to a single thruster. The transfer or dual feed mechanism introduces
additional complexity and brings with it additional scrutiny from regulators but provided this
is properly managed the design is a useful way of enhancing the post-failure capability of the
simple two-way split.
The four-way split is also popular and is better suited to the thruster arrangement of semi-
submersibles than monohulls. There are a large number of variations on this design. In some
arrangements the power distribution is arranged to ensure that there is always at least one
thruster operating in each corner following a single failure. In other arrangements the design
accepts loss of thrust in one corner. Figure 78 shows one possible variation on this theme.
G G G G
G G G G

Figure 78 Four-way split for semi-submersible
In all designs it can be beneficial to closely associate the power supplies for thruster auxiliary
systems with the supply to the main motor itself. Some designers choose to derive the
auxiliary supply from the drive transformer by way of a fused isolator while others provide a
dedicated step down transformer from the main power system. The latter arrangement may
be preferred if the drive auxiliaries must be energised from the main power system before
starting the drive as the drive circuit breaker will be closed by the drive itself after
pre-charging.
4.2.2 Innovations
There have been several innovative designs for improving redundancy and fault tolerance in
diesel electric systems, some dating back to the early 1980s, before the DP rules and
guidelines were well established. Others have been made possible by advances in power
electronic converters and their control systems. It is interesting that the impetus for many

92 IMCA M 206
innovations originates with vessel owners seeking to improve reliability. A few of these
innovative designs are discussed in the sections that follow.
4.2.3 Split Reactor
It has long been recognised that one of the greatest challenges in the design of fault tolerant
diesel electric power systems is to ensure that consumers can survive the severe voltage dip
associated with clearing a short circuit fault anywhere on the main power distribution system.
In the early 1980s one of the classification societies developed a power plant design for low
voltage diesel electric systems which allowed all generators to supply power to all thrusters
while at the same time protecting the surviving thruster motors against the severe voltage dip
associated with a fault on one of the two main switchboards. Figure 79 shows the overall
single line diagram.
In a conventional diesel electric power system with a two-way split operating as a common
power system a short circuit anywhere on the main power distribution system can pull the
system voltage down to zero until it is cleared by the over current protection. The resulting
disruption can cause thruster and auxiliary system motors to trip. In the split reactor system
each generator supplies both main switchboards by way of a split reactor and there is no
direct connection between the two switchboards as in a conventional power plant. If a fault
occurs on one main switchboard (point A) it will cause the voltage at the generator terminals
to fall, but the voltage drop across the split reactor is coupled across to the healthy
switchboard and used to maintain the system voltage until the fault is cleared. All generators
are left connected to the healthy switchboard. The split reactor system also significantly
reduces the magnitude of the fault current allowing all generators to be connected to one
switchboard and reducing the need for a switchboard with a high short circuit current rating.
In conventional low voltage power systems it is sometimes necessary to open the bustie
when the fourth generator connects to limit the fault current that can be experienced by
each main switchboard. During normal operation the split reactor system has no core losses
provided the thruster loads are reasonably well balanced.
G
G
G
G
VOLT RISE VOLT DROP
GENERATOR
TERMINAL
VOLTAGE FALLS
A
FAULT CURRENT
BUS VOLTAGE
MAINTAINED
VOLT RISE VOLT DROP
VOLT RISE VOLT DROP
VOLT RISE VOLT DROP

Figure 79 Split reactor power plant

IMCA M 206 93
4.2.4 Phase Shifted Power Systems
One marine power system provider developed a method to reduce the electrical losses
(wasted power) associated with the phase shifting transformers required for harmonic
cancellation in propulsion systems utilising variable speed drives. Figure 80 illustrates the
general principle of operation. Rather than provide a phase shifting transformer for each
thruster drive, two phase shifting transformers (one for each half of the power plant) are
used to phase shift the generators connected to the two main busses with respect to each
other. One of the two rectifiers for each thruster drive is connected to each bus and thus
the drives draw power directly from the generators with the correct phase shift for harmonic
cancellation.
G
690V
G G G
690V
PHASE SHIFTING
TRANSFORMER
690V
690V
PHASE SHIFTING
TRANSFORMER

Figure 80 Phase shifted power system
4.2.5 Grid Interconnector
One vessel owners design initiative led to the development of a new type of power system
for large vessels which can be applied to DP Class 2 and DP Class 3 designs. In this design,
grid interconnector technology is used to provide each thruster variable speed drive with a
continuous dual feed from both main switchboards as shown in Figure 81. Each thruster is
designed to be independent of the others and thus the worst case failure is limited to loss of
a single thruster. In addition to supplying the thruster drives from a dual source, the thruster
auxiliaries are also supplied in the way shown in Figure 82. The thruster auxiliary system can
be powered from a large battery bank for a short period of time and from the emergency
generator if required. The battery bank allows the thrusters to be made ready for DP
independently of the generators and thus thruster and generator recovery become parallel
activities, rather than sequential ones, saving time in the recovery sequence.
T 1 FWD
STARBOARD
PORT
T 2
T 3
T 5
T 6
G2 G3 G1
G5 G6 G4
VESSEL SERVICES AFT T 4

Figure 81 Grid interconnector based power system

94 IMCA M 206
480V BUS E 11 kV BUS B
AUXILIARY CONVERTER
1 kVdc
MAIN CONVERTER
THRUSTER AUXILIARY
PUMPS AND FANS
1.8 kVdc
THRUSTER
11 kV BUS A

Figure 82 Power supplies to thrusters and their auxiliary systems
4.2.6 Practical Implementation in Conventional Diesel Electric Designs
The design process for the DP system of a DP Class 2 or DP Class 3 vessel should begin with
an estimation of the environmental conditions in which the vessel is expected to hold
position and heading. In some cases this may be expressed as a percentage uptime at a
particular location based on historical data regarding conditions at that location. The
required thruster force defines the required post-failure DP capability. From this
information the designer develops the redundancy concept to ensure that sufficient
propulsion machinery is available to develop the required thrust after the worst case failure
has occurred. A suitable margin should be included for deteriorating weather conditions,
giving due consideration to the length of time necessary to make operations safe. Power may
also be required after the worst case failure for non-propulsion related machinery such as
drilling equipment, cranes and pipelaying equipment.
Even within conventional diesel electric designs there is considerable variation in the practical
implementation of redundancy concepts. Four examples are used to illustrate some of the
features found on vessel of different types.
Figure 83 shows the single line power system diagram for a small DP Class 2 power plant as
might be fitted to a platform supply vessel or an anchor handling tug. For this size of vessel
the power generation system may be 440V or 690V and power is provided from a
combination of shaft generators fitted to the large main engines and auxiliary generators.
The two controllable pitch propellers at the stern are driven directly by the main engines.
Four controllable pitch tunnel thrusters provide athwartships thrust and these are driven by
the diesel electric plant. The main engines and controllable pitch propellers provide high
transit speed and bollard pull. For DP operations the tunnel thrusters are combined with the
main propellers. In some designs the rudders are also interfaced to the DP control system.
The redundancy concept is essentially a two-way split and pumps and fans for auxiliary
systems are powered directly off the main 440V switchboard. In vessels with 690V power
generation larger motors may be powered from the 690V system and smaller consumers
from a 440V distribution system. It is possible to design such a power plant to be operated
as a common power system or as two independent power systems but more complex
protection is required for the common power system design.


I
M
C
A

M

2
0
6


9
5


G2 G1 SG2 SG1
220V
24Vdc
220V
440V
STARTING
AIR COMP
2
STARTING
AIR COMP
1
DP
UPS 2
DP
UPS 1
EMERGENCY
SERVICES
HPU
CPP6
T2
HPU
T2
T4
HPU
T4
FO 2 FW 2 SW 2 VENT 2 FO 1 FW 1 SW 1 VENT 1
HPU
T1
T1
HPU
T3
T3
HPU
CPP5
EMERGENCY
EG
440V
220V
24Vdc
440V
CPP 5
CPP 6
DG2
DG1
SG
2
SG
1
NC
NC
T4 T3 T2 T1

Figure 83 Combined mechanical and electric DP power plant

96 IMCA M 206
The power plant shown in Figure 84 is typical of all electric DP vessels such as diving
support vessels, ROV vessels or smaller pipelayers. The thruster arrangement includes two
tunnel thrusters and a retractable azimuth thruster at the bow but larger vessels may have
two or three retractable azimuth thrusters forward in addition to the tunnel thrusters.
Tunnel thrusters offer advantages for vessels that make frequent port calls as part of normal
operations. The aft thrusters are often sized for high transit speed and are larger than
required for DP. Such designs typically have high voltage power systems such as 6.6kV and
marine auxiliary systems are supplied at lower voltage such as 480V. 690V designs may be
possible for installations less than 10MW and typically have four generators rather than six.
In such designs it may be necessary to open the busties when all four generators are
connected.
Thrusters are typically fixed pitch and driven by variable speed drives but some designers
choose to specify controllable pitch propellers for the propulsion thrusters and vary both the
speed and pitch to achieve the best performance in DP and transit conditions.
The redundancy concept is a two-way split but the forward azimuthing thruster and its
auxiliaries can be powered from either power system. Whether the transferable thruster can
be considered to contribute to the vessels post-failure DP capability may depend on the DP
notation being sought and the classification society and designers are advised to clarify these
points early in the design phase. Some designers choose to add a third retractable azimuthing
thruster at the stern. Provided this is also arranged with a transferable power supply it can
be used to maintain the fault tolerance of the power plant in the event that one of the stern
thrusters fails.
The power plant is typically divided into two engine rooms and two switchboard rooms even
for DP Class 2 designs. DP Class 3 versions of this design are possible, provided additional
isolation (circuit breakers or fast acting fuses) is provided for cables connecting redundant
elements. Voltage dip ride through must also be carefully considered.
Normal power supplies for UPSs and DC power supplies are provided from the main power
distribution system in a manner which matches the port and starboard split in the redundancy
concept. All battery systems have a backup supply from the emergency switchboard by way
of an auto changeover with suitable protection to prevent a fault in the UPS being transferred
from one supply to the other. This feature provides some protection against poor battery
endurance as the emergency generator should connect to re-supply the UPS or DC power
supply in 45s or less. It is also useful to be able to charge batteries if the main power system
is unavailable but it is good practise to make blackout recovery of the main power system
independent of the emergency generator.



I
M
C
A

M

2
0
6


9
7


G3 G2 G1 G5 G4 G6
T1
TUNNEL
THRUSTER
PORT 6.6kV
NC NC
STBD 6.6kV
NC NC
STBD 480V
T4
PROPULSION
AZIMUTH THRUSTER
T2
TUNNEL
THRUSTER
PORT 480V
T5
PROPULSION
AZIMUTH THRUSTER
T3
RETRACTABLE
THRUSTER
T3 MCC
EMERGENCY 480V
STARTING
AIR COMP
2
STARTING
AIR COMP
1
EG
PORT 220V
T2 MCC
FW PUMP
3
FW PUMP
4
SW PUMP
3
T5 MCC
SW PUMP
4
VENTILATION 110Vdc
2
24Vdc
2
VMS
UPS 2
DP
UPS 2
LIGHTING
T1 MCC
FW PUMP
2
FW PUMP
1
SW PUMP
2
T4 MCC
SW PUMP
1
VENTILATION 110Vdc
1
24Vdc
1
STBD 220V
VMS
UPS 1
DP
UPS 1
LIGHTING LIGHTING
STARTING
AIR COMP
3
PORT ER
SW HEAT
PRELUBE
STBD ER
SW HEAT
PRELUBE
EMERGENCY
SERVICES
5
4
3
2 1
DG5
DG4
DG3
DG2
DG1
FO PUMP
2
FO PUMP
1
HPU
FW
LO
HPU
FW
LO
HPU
FW
LO
DP CONTROL
CONSUMERS
OS
FS
DP CONTROL
CONSUMERS
OS
FS
DG6
LIGHTING

Figure 84 DP power plant with two-way split

98 IMCA M 206
Figure 85 shows a multi-way split with a six-way split in the HV power system and a four-way
split in the LV power system. This arrangement is favoured by some designers who want to
reduce the impact of the worst case failure under DP Class 2 failure criteria to a single engine
and thruster without resorting to features such as transferable generator or thrusters.
Typically, the four-way split is used to provide redundant power supplies to duty standby
pumps and marine auxiliary systems may be provided as a two-way split exploiting the
exemption of pipe work failure under DP Class 2 criteria. In this case care must be taken not
to fall foul of those classification societies that require separation of FW cooling, fuel and
pneumatic supplies for certain DP Class 2 notations.
Sometimes this power system arrangement is used in DP Class 3 designs where the vessel
owner wants to define two post-failure DP capabilities, one for DP Class 3 failure criteria and
the other for DP Class 2 failure criteria. The justification for this is that the vessel may only
carry out the most critical DP operations in good weather and therefore does not require a
very high level of post-failure DP capability. The DP Class 3 redundancy concept may
therefore be designed as a simple two-way split based on the physical separation provided by
two engine rooms and two switchboard rooms. Less critical operations may be carried out
in higher environmental conditions. In this case the vessel operates within an environmental
envelope defined by the DP Class 2 redundancy concept which is loss of one generator and
one thruster. In other words, in the increased environmental conditions, the vessel can
withstand any single failure under DP Class 2 failure criteria without losing position or
heading.
Although the idea that a DP vessel can operate in more than one power plant configuration is
not new there is increasing awareness amongst classification societies that some
configurations may not be fully fault tolerant. At least one classifications society now states
that vessels are considered to comply with the requirements of their DP rules only when
operated in a DP system configuration analysed in the approved FMEA. Until recently there
was no formal recognition of a dual DP equipment class and vessels were assigned a notation
equivalent to IMO DP Equipment Class 1, 2 or 3. However, at least one classification society
now offers a dual DP2 and DP3 notation provided the FMEA and proving trials fully address
both DP Class 2 and DP Class 3 failure criteria in all defined operating configurations.


I
M
C
A

M

2
0
6


9
9


G4 G3
6.6kV STBD
G2 G1
440V STBD
6.6kV PORT
G6 G5
EMERGENCY
EG
230V
6
4
3 1
DG5
DG4
DG3
DG2
DG1
2
5
T2 T3 T6 T4 T5 T1
440V
DG6
440V PORT
230V
110Vdc 24Vdc 110Vdc 24Vdc 110Vdc 24Vdc
FWPUMP
4
SWPUMP
4
FO PUMP
4
FWPUMP
3
SWPUMP
3
FO PUMP
3
FWPUMP
1
SWPUMP
1
FO PUMP
1
FWPUMP
2
SWPUMP
2
FO PUMP
2
STARTING
AIR
110Vdc 24Vdc
VMS
UPS 3
DP
UPS 3
VMS
UPS 1
DP
UPS 1
VMS
UPS 2
DP
UPS 2

Figure 85 DP power plant with multi-way split

100 IMCA M 206
The power plant shown in Figure 86 is typical of larger DP vessels such as drill ships, heavy lift
vessels or large pipelayers. Many designers opt for a three- or four-way split in order to
reduce the impact of the worst case failure. This helps to reduce thruster and engine size.
Tunnel thrusters and retractable units may offer little advantage for vessels which operate in
one location for long periods of time.
Vessels with a large installed power typically have a 6.6kV or 11kV power systems. The
three-way split may be carried right through the redundancy concept for all propulsion and
DP related consumers but other arrangements may be made for non-DP or propulsion
related consumers such as heating, lighting, accommodation HVAC and similar vessel
services.
This design can be made to operate as a closed ring using differential protection to identify
and disconnect only the faulty bus section. This allows great flexibility in the assignment of
generators. The power plant can also be designed to operate as three independent power
systems. The power system will generally be split into three engine and switchboard rooms
even for DP Class 2 versions.
In this particular design the power distribution systems are carefully split to ensure that each
power system is as independent as possible and there are as few common point as possible
connecting them. The separation is carried right through all distribution voltage levels from
11kV down through 480V, UPSs and DC supplies. This level of physical separation limits the
opportunity for faults to propagate from one redundant system to another.
The three-way split also ensures that the vessel is fully fault tolerant with a reduced
environmental envelope even with one power system unavailable. This can be a useful
feature to allow continued operation while maintenance or major repair work is carried out.
Providing power to large non DP consumers such as drilling systems, pipelaying systems and
cranes is one of the challenges facing designers of multi-split power systems. Although the
design offers a high degree of fault tolerance, each individual power system tends to be
smaller than the equivalent two-way split. Not only is the power plant subdivided into
smaller sections but the overall generating capacity can be significantly less than an equivalent
design based on an two-way split. This occurs because the worst case failure results in the
loss of fewer generators so they can be smaller for the same post-failure DP capability. To
overcome this problem it may be necessary to provide large consumers with a feed from
each power system. It may also be necessary to consider means to ensure that power is
shared equally amongst power systems or in proportion to their capacity. Further
complications may be encountered if the power plant must be operated as several
independent power systems, particularly in relation to the operation of harmonic cancellation
and regenerating power to each power system.


I
M
C
A

M

2
0
6


1
0
1


11kV
G4 G3
MCC
TMCC TMCC
MCC
11kV
G2 G1
MCC
TMCC TMCC
MCC
480V
NON PROPULSION
SERVICES
480V
NON PROPULSION
SERVICES
480V
11kV
G6 G5
MCC
TMCC TMCC
MCC
480V
220V 220V 220V
110Vdc 24Vdc
VMS
UPS
EMERGENCY
EG
110Vdc 24Vdc
VMS
UPS
110Vdc 24Vdc
VMS
UPS
220V 220V 220V
DP
UPS B
DP
CONSUMERS
DP
UPS C
DP
CONSUMERS
DP
UPS A
DP
CONSUMERS
480V
6
4
3
1
DG5
DG4
DG3
DG2
DG1
2 5
T1 T3 T6 T4 T2 T5
480V
AIR
VENT
SW
FW
FO
FW
VENT
UPS
HPU FW
VENT
UPS
HPU FW
VENT
UPS
HPU FW
VENT
UPS
HPU FW
VENT
UPS
HPU FW
VENT
UPS
HPU
DG6
AIR
VENT
AIR
VENT
AIR
VENT
AIR
VENT
AIR
VENT
SW
FW
FO
SW
FW
FO
SW
FW
FO
SW
FW
FO
SW
FW
FO

Figure 86 Power plant with three-way split for large DP vessel

102 IMCA M 206
4.3 Operational Configuration
4.3.1 Choice of Operational Configuration
All DP Class 2 and DP Class 3 power plants are intended to be fully fault tolerant when
operated in one of the configurations analysed in the vessels approved DP FMEA.
Conventional diesel electric power plants can usually be operated in one of two
configurations. Either as a single common power system capable of automatic subdivision in
response to a fault or as two or more independent power systems. These modes of
operation are sometimes referred to as busties closed and busties open respectively.
Much debate in the DP community has centred on the relative advantages and disadvantages
of these two modes of operation but one thing is certain, neither configuration guarantees
station keeping integrity on its own and the whole issue of DP safety and reliability is far
more complex than choosing one power plant configuration or the other.
The issue is not whether one power plant configuration is safer than another but rather
whether each chosen configuration is adequately safe given the consequences of loss of
position, i.e. an absolute test rather than a relative one. IMCA M 181 Analysis of station
keeping incident data 1994-2003 presents some interesting conclusions of the subject of
open and closed busties for those readers who wish to know more.
4.3.2 Common Power System
The greater the number of generators connected to a power system, the more robust is that
power system because it becomes increasingly difficult for a single generator to vary the
power system voltage and frequency. Common power systems allow greater flexibility in
generator utilisation which translates to lower fuel consumption, less pollution and reduced
maintenance.
Marine power systems are relatively small and DP vessels operating with a common power
system are heavily dependent on a wide range of protective functions for fault tolerance.
They are also heavily dependent on the voltage dip ride through capability of all generators
and consumers connected to that power system. Neither of these features is fully tested at
sea trials and therefore much depends on the quality of the power system studies on which
the protection schemes are based and on the commissioning of the protective functions
themselves. At least one of the major classification societies does carry out live short circuit
testing of power systems for redundant propulsion systems and such tests do reveal
inadequate protection schemes and poor fault ride through capability.
4.3.3 Independent Power Systems
When diesel electric systems are designed to operate as two or more independent power
systems each power system may consist of only one or two generators and the ratio of
generator rating to system load is greater than in the common power system case. That is to
say the system load is smaller as it is divided amongst several independent power systems.
Thus it is easier for a faulty generator to force healthy generators to trip.
In order to run all the thrusters there must be at least one generator running on each power
system. As such, in light load conditions the reliability of the power system, and thus the
thrusters, is heavily influenced by the reliability of a single diesel engine which is relatively low.
Therefore, there may be a more frequent occurrence of partial blackout and loss of
thrusters. Provided the vessel is being operated within its post-failure DP capability for that
power plant configuration, there should be no significant position excursion but this will
increase the number of times that the remaining generators and thrusters are called upon to
compensate rapidly for the loss of one power system. It can be argued that, in practical
terms:
common power systems are vulnerable to the hidden failure of protective functions;

IMCA M 206 103
multiple independent power systems are vulnerable to failure of redundant elements to
achieve their rated capacity. This includes equipment such as cooling systems as well as
thrusters and generators.
4.4 Transferable and Dual Fed Consumers
4.4.1 General
Providing means by which a thruster or generator can connect to more than one main
switchboard is a way of reducing the permanent impact of the vessels worst case failure
without further subdivision of the power plant. However, the redundancy concept and
therefore the vessels post-failure DP capability is dependent upon successfully restarting the
transferable generator or thruster on the healthy side of the power plant. The transfer needs
to be carried out fairly rapidly to limit the position excursion and times of ten seconds or less
are often specified. Note that the views of classification societies differ on whether or not
transferable generators and thrusters can be considered to contribute to redundancy. It is
highly advisable to consult class at an early point in the development of the basic design if
considering having transferable or dual fed equipment as part of the redundancy concept.
4.4.2 Generators
Some classification societies will accept transferable generators as contributing to the vessels
post-failure DP capability and others may not accept such features or only for certain
notations. However, even those classification societies that require redundancy to be based
on running machinery will generally allow such features to be included in the design even
though they are not part of the redundancy concept. Some vessel designers add transferable
generators as a means of recovering power for process activities, such as drilling or
pipelaying, but not for station keeping.
4.4.3 Thrusters
In the past ten years there have been significant changes in attitudes towards transferable
thrusters. Originally, all the major classification societies were prepared to accept thrusters
with automatic changeover as contributing to the vessels post-failure DP capability. Opinions
then changed and it became difficult to have any type of changeover thruster accepted by
some classification societies. However, designs which transferred the thruster supply without
the thruster stopping or dropping out of DP were viewed more favourably than others.
More recently opinions have changed again with the availability of power electronic
changeovers and true dual feed arrangements such that these types of features can be
accepted as contributing to post-failure capability on DP Class 2 and DP Class 3 designs.
However, particular attention must be paid to eliminating the possibility of a fault in one
power system or the thruster being transferred to the other power system. It is also very
important to reduce the risk of hidden failures which may prevent the transferable thruster
connecting successfully to its alternate source of power. Careful consideration should be
given to alarms and monitoring, internal diagnostics and test facilities to give a high degree of
confidence that the changeover will operate successfully on demand.
4.4.4 Essential Services Switchboards
Some vessel designers elect to provide motor control centres for essential engine room
services with two sources of power. The normal supply is typically the LV distribution
associated with the generators in that engine room. The alternate source of power is usually
the emergency switchboard. Note that this arrangement is only practical if the vessel has a
large emergency generator. The essential services switchboards for each engine room will
typically supply such things as:
electric fuel pumps;
FW cooling pumps;
SW pumps;
starting air compressors;

104 IMCA M 206
combustion air fans;
engine jacket water pre-heaters;
pre-lub oil pumps.
Note that some classification societies will not accept that the essential services switch
automatically to the emergency switchboard as this may compromise the availability of the
emergency generator. Therefore it is often necessary to ensure that blackout recovery is
possible with none of the essential services available, at least in the short-term. Making
blackout recovery independent of the emergency generator is good practice in any case.
Some designers have chosen to install an essential services generator for this purpose rather
than use the emergency generator, and then seek exemption from having an emergency
generator, on the basis of having a fully redundant power system. Note that some fairly
onerous conditions may be applied to such an exemption.
4.4.5 Transfer of Fault
In all cases where one power system or one control system is connected to another by
shared equipment such as a thruster or generator there is a risk of transferring a fault from
one system to the other. In general, transferable or dual fed equipment must be provided
with a comprehensive range of protective functions to prevent or limit the effects of faults
that could transfer in this way.
4.5 Battery Systems
4.5.1 General
IMCA M 196 Guidance on the design, selection, installation and use of uninterruptible power
supplies onboard vessels provides information on how UPSs and DC power supplies can be
integrated into the redundancy concept.
4.5.2 DC Supplies for Switchboard and Generator Controls
DC supplies with battery backup are used for switchboard controls, governors, AVRs,
thruster control and navigation equipment etc. Typical voltages are 125Vdc, 110Vdc and
24Vdc. In the absence of any other requirement, the battery endurance for generator and
switchboard related equipment is typically sized to allow 30 minutes operation, but it is
prudent to size battery banks with a generous margin to allow for ageing.
As the power consumption of a switchboard control system increases greatly when the
circuit breakers are being opened and closed it may be appropriate to consider the need to
manoeuvre the switchboards during blackout in addition to the static control load. It may
also be prudent to provide separate battery supplies for spring winders and control supplies
to prevent a voltage dip associated with multiple simultaneous operations of spring winders
adversely affecting the control voltage.
Loads such as space heating should be separated from control supplies.
It is beneficial to provide all battery systems with a normal supply from the main power
distribution and a backup supply from the emergency power distribution.
4.5.3 UPSs
Modern DP vessels make extensive use of UPSs to supply equipment such as:
DP control system;
vessel management system operator stations and field stations;
network hubs and switches;
power management system;
thruster controls;

IMCA M 206 105
pre-charge and ride-through supplies for UPSs.
Figure 87 shows the schematic of a double conversion UPS. In this design the load is
normally supplied by way of the battery charger and inverter. The bypass is only used for
battery maintenance purposes or if the inverter fails. This is the most popular UPS for
marine applications and the type recommended for DP vessels. Other designs are available
such as line interactive types. In this design the load is normally fed from the bypass and is
only switched to the inverter when the main supply fails. There have been cases of this type
of UPS switching to the inverter on detection of low power quality which is common in some
vessel power plants. This happened so frequently that the battery was almost depleted when
needed.
OUTPUT
CIRCUIT
BREAKER
MAIN AC
INPUT
LOAD
LINE
CONDITIONING
BATTERY
INVERTER
BYPASS
CONTROL
CCT
BATTERY
CHARGER
FILTER
SENSE LINE

Figure 87 Double conversion or online UPS
It is beneficial to provide all UPSs with a normal supply from the main power distribution and
a backup supply from the emergency power distribution.
The ability of UPSs to supply sufficient fault current to operate over current protection
selectively needs careful consideration in any application where redundancy depends upon
the selectivity of circuit breakers or fuses.
There are several strategies for providing UPS power to the various consumers and these are
discussed in detail in IMCA M 196, but the two extremes are:
two large centralised UPSs;
distributed UPSs for each piece of equipment.
Both methods have their advantages and disadvantages but it can be difficult to overcome
problems associated with transfer of fault and poor selectivity in the centralised arrangement,
particularly for DP Class 3 designs.
4.6 Separation of Redundant Elements for Fire and Flood
4.6.1 Requirements for Physical Separation
Although the requirements for physical separation of redundant equipment are focused on
DP Class 3 designs there are requirements to consider the possibility of mechanical damage
and fire in DP Class 2 designs. Even though DP Class 2 designs may have a single engine
room, common cable routes serving redundant equipment may not be accepted if they pass
through high risk areas. In general, it is considered good practice to separate cables for
thrusters and controls systems as far as practical, even in DP Class 2 designs.
4.6.2 Cable and Pipe Routes
Physical separation of cable and pipe routes serving redundant equipment is required for all
DP Class 3 designs. Rules and guidelines vary on what level of separation should be provided.
IMO MSC 645 requires that cables and pipes serving equipment intended to provide

106 IMCA M 206
redundancy should be separated by bulkheads of A60 classification. Cables may pass through
the same compartment in ducts of A60 rating if the only fire risk within the duct comes from
the cables themselves; similarly for pipe work. Most classification societies will accept
crossover pipe work connecting redundant system in DP Class 3 designs, provided it can be
isolated on either side of the bulkheads providing physical separation. Without such isolation
facilities it may be necessary to allow a compartment to continue to flood so that cooling of
other systems can continue (in the case of seawater cooling systems).
4.6.3 Fire Subdivisions
Some classification societies only require that redundant equipment and cables are in separate
compartments, no mention is made of the rating of bulkheads or deck heads. Other
classification societies require A60 classification but will accept two A0 bulkheads as satisfying
the requirements for physical separation, provided the space between them is a low fire risk.
4.6.4 Watertight Compartments
IMO MSC 645 requires physical separation of equipment installed below the operational
water line. Equipment intended to provide redundancy must be located in watertight
compartments, complying with requirements for fire protection. Some classification societies
make reference to the damaged water line and at least one classification society requires that
every thruster is in its own A60, watertight compartment. Others may accept that thrusters
are grouped in compartments reflecting the overall split in the redundancy concept.
Watertight fire dampers may be arranged to prevent down flooding and the failure modes of
these devices need to be considered in the redundancy concept.

IMCA M 206 107
5 Thrusters, Drives and Controls
5.1 General Propulsion Principles
5.1.1 The Origins of the Propeller
The Archimedes screw was used to move water as part of irrigation schemes in the 3rd
century BC, but the first use of a screw as a means of propulsion occurs much later in
history:
James Watt said Have you considered a spiral oar? (1770);
Joseph Bramah patented a screw propeller but never developed it practically (1785);
Joseph Ressel is credited as the inventor by Austria (patent 1827);
Richard Trevithick described one in an 1815 patent;
John Swan is heralded as the practical inventor, using one in 1824.
5.1.2 Relationship between Propeller Power, Thrust and RPM
Sir Isaac Newtons third law of motion states that each action has an equal and opposite
reaction. A propeller develops thrust by accelerating water to a given velocity. The water
velocity V is proportional to the speed of the propeller (RPM).
From
2
2
1
MV F it can be said that:
2
RPM Thrust or
2
1
Thrust RPM
The power to develop this thrust follows the relationship:
Velocity Thrust Power
Therefore:
3
RPM Power or
2
3
Thrust Power
Equation 9 Relationship between power and thrust
Figure 88 shows the relationship between power and thrust for an actual azimuthing thruster
in bollard pull conditions. The figure was plotted using data supplied by a manufacturer.
DP thrusters only operate in bollard pull conditions when the vessel speed is zero. When
the current is non-zero there may be water flow into the thruster even though the vessel is
stationary.

108 IMCA M 206
0 10 20 30 40 50 60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
Power Against Thrust Typical FP Thruster (Bollard Pull)
P
e
r
c
e
n
t
a
g
e

P
o
w
e
r

(
%
)
Percentage Thrust (%)

Figure 88 Relationship between power and thrust
5.2 Thrusters
5.2.1 Thruster Failure Modes
Some DP rules and guidelines have specific requirements that thrusters should fail safe and
not develop uncontrolled thrust or change direction with thrust applied as the result of a
single failure. Fail safe conditions are generally considered to be:
fail as set;
fail to zero thrust;
motor stop;
uncontrolled change in thrust direction may be accepted provided thrust goes to zero.
Fail as set may not be a good choice in some situations.
5.2.2 Thruster Types
There are several different types of thruster:
propeller with high lift rudder;
gill jet;
Voith Schneider;
tunnel thruster;
azimuth thruster (several forms).
Propellers can be of fixed pitch or variable (controllable) pitch.
Propeller: This is a component part of many thrusters as well as the most common form of
propulsion. The propeller can be of fixed pitch or variable pitch. The pitch of a propeller is
the theoretical distance moved through the water for one revolution, but due to slippage this
is never achieved. One way to visualise this is to consider the axial distance moved when a
wood screw is turned through one revolution in a piece of wood. Propeller efficiency is an
important consideration as if the propeller is not correctly matched to the vessel then it will
never perform to expectations. This is not only an important consideration for operational
costs but also for the environmental emissions.

IMCA M 206 109
The high lift rudder can be connected to a DP control system and may be accepted as
contributing to athwart-ships thrust in certain conditions.
Gill jet: This system is not as common now as it was in the past. The thruster consists of
an axial flow pump with vertical shaft axis delivering water downwards across a grill with
angled plates (gills) in the bottom of the vessel to direct the thrust in the required direction.
The gill plate is circular and can be rotated through 360 to provide a multidirectional thrust
capability.
Voith Schneider unit: This type of unit is capable of thrust delivery in any direction.
When used as a means of propulsion it does not need a rudder. The blades are attached at
right angles to the rotor casing and rotate around a vertical axis. Each blade performs an
oscillatory motion around its own axis. This is superimposed on the uniform rotary action of
the entire unit. When the unit is fitted in the hull, only the blades protrude from the hull, as
shown in Figure 89.

ROTOR
BLADES

Figure 89 Voith Schneider propeller
Tunnel thruster: The tunnel thruster requires a tubular water passage running athwart-
ships with the unit placed at its centre. This allows for the thrust to be directed in either
port or starboard directions by reversing propeller pitch or direction of rotation. The
limitations of this type of thruster are the length of tunnel in which it is situated and the
distance it is located from the bow or stern. The longer the tunnel the greater the possibility
of cavitation at high loads as the water flow may become restricted. The further from the
bow or stern the less the turning moment created about the vessels centre of rotation.
Tunnel thrusters located at the stern may also be susceptible to aeration of the water caused
by the main drive propellers. There are no protrusions from the hull when tunnel thrusters
are used.
Rim drive thrusters are a relatively recent innovation which essentially remove the need for a
gearbox as the rotating element is the propeller which sits inside the stator of the motor.
This has the advantage of reducing the central body of the propeller, thus aiding the flow of
water.
MOTOR
GEARBOX
TUNNEL
PROPELLER

Figure 90 Tunnel thruster

110 IMCA M 206
Azimuth thrusters: The azimuth thruster is mechanically similar to the tunnel thrusters, it
has the advantage however of being able to direct the thrust in any direction as opposed to
port and starboard only. It also operates in open water which has fewer problems in relation
to the dynamics of flow as compared to the tunnel thruster.
MOTOR
GEARBOX
NOZZLE
PROPELLER
AZIMUTHING
GEAR

Figure 91 Azimuthing thruster
Propulsion thruster: This is used in the same manner as the conventional shafted
propeller system, with steering being achieved by rotating the thruster rather than operating
a rudder. It also forms part of the station keeping system when operating in DP. On some
types of vessel all the thrusters are of this type and are designed to be removed without the
need to put the vessel in drydock. Class will apply elements of steering gear rules to
designated propulsion thrusters.
Retractable azimuth thruster: This is similar to the propulsion thruster but it can be
withdrawn into the hull of the vessel in order that it does not create extra drag while the
vessel is in transit. For short transit distances, the thrusters may be left deployed and under
power. For long transit distances, the increase in speed achieved by their use is not justified
normally due to the high additional operating cost incurred.
Combined retractable thruster and tunnel thruster: Theoretically, this provides the
advantages of both types but as with any multifunctional system it is a compromise which may
be suitable for some situations but not for all. As the hull has less material the possibility of
structural deformation as a result of the forces developed increases, therefore the hull has to
be considerably strengthened to compensate for this loss of strength. The increase in weight
created may be detrimental to the vessel or the thruster power may need to be decreased to
accommodate it adequately. It also has to be drawn further into the vessel thus increasing
the use of internal space.
Contra rotating azimuth thruster: Thrusters with contra-rotating propellers offer
higher efficiencies of between 10-15% because the aft propeller regains some of the energy
losses in the stream as well as rotational losses. Contra-rotating propeller thrusters also
have low noise and vibration, and for the same power have propeller diameters 20% smaller
than single screw units, giving a shallower draught. The unit requires a variable speed drive as
there is no option for a CP propeller. There are also more complex sealing and thrust
containment systems to consider.
Podded azimuth thruster: The podded drive provides the means to deliver greater
power than previously possible with a geared azimuth thruster by eliminating the gear train in
the thrusters, as shown in Figure 92. The shaft of the thruster is also the motor rotor with
the pod casing being the stator. The seal arrangement becomes much more critical as there
is now the possibility of electrical failures occurring as a result of seawater ingress. A bilge
arrangement is provided to remove any leakage into the pod itself and assess the rate of
leakage. Power ratings are of the order of 2MW to 25MW.

IMCA M 206 111
MOTOR
PROPELLER
SHAFT
POD
BRAKE
SLIP RING

Figure 92 Podded thruster
Azipull thrusters: This appears to be a back to front azimuth thruster but there are
advantages in this design. A pulling propeller (CPP or FPP) is mounted ahead of the leg,
which is a streamlined unit incorporating the gear house and a lower fin. The leg/housing/fin
combination recovers swirl energy from the propeller slipstream which would normally be
wasted, converting it into additional forward thrust. At the same time the underwater unit
has more steering effect than a conventional azimuth thruster, improving the steerability of
many hull forms.
Portable thruster: At least one manufacturer now offers hydraulically driven swing down
thrusters which are largely independent in terms of power and control and can be added to a
dumb barge to provide a DP capability with relatively little effort compared to a conventional
unit.
5.2.3 Choice of Thruster
In any DP Class 2 or DP Class 3 new building or conversion project, the choice of which
thrusters to use is often made at an early stage in the basic design process due to the long
manufacturing lead times for such units. In addition to lead time, there are many other
factors to consider when choosing a thruster for a particular application such as:
thruster type tunnel or azimuthing;
thrust capability;
physical size headroom under deck head etc.;
fixed or retractable;
variable speed, variable pitch or combinatory;
electrical drive, direct diesel drive;
reliability;
maintainability;
availability of service engineers.
One of the most important points to consider is:
How will the choice of thruster influence the development of the redundancy concept?
The choice of thruster type will significantly influence the redundancy concept and it is
important to ensure that the redundancy concept incorporates the necessary features to
support that particular choice.
5.2.4 Thruster Power
DP Class 2 and DP Class 3 vessels should have thrusters installed in number and power
rating such that they can maintain position and heading following the worst case failure.
The surviving thrusters must be able to generate the required surge, sway and yaw forces.
The surviving generators must be capable of supplying the power required by the thrusters.

112 IMCA M 206
The vessels post-failure DP capability is determined by these factors. Stern thrusters may be
sized for transit speed and may operate at a fraction of their rating on DP.
5.2.5 Physical Constraints
Fixed pitch thrusters driven by variable speed AC drives are very popular but it is not always
appreciated how much space and weight can be taken up by the converter and its related
support equipment, such as drive cabinets, phase shifting transformers, de-ionized cooling
water units, UPSs, pre-charge units, etc. There will also be a need to provide connections to
FW cooling systems, HVAC and electrical supplies for all these units.
Although the variable pitch propeller may have some perceived disadvantages in terms of
increased maintenance requirements and lower reliability due to its mechanical complexity, it
can be packaged into a very compact arrangement at low and medium power levels and
requires very little in the way of ancillary equipment and support services. This might be an
important consideration in small and medium sized DP vessels.
5.2.6 Low Load Performance and Related Issues
The advantages of the fixed pitch, variable speed thruster are its mechanical simplicity and
low power consumption at low propeller speed. Many DP vessels spend only a fraction of
their working life operating in conditions close to their maximum post-failure capability and
therefore thrust demand levels can be very low much of the time. The result is that the
vessel has to operate with a few lightly loaded generators online which can be an
uncomfortable condition both in terms of power plant stability and running conditions for
diesel engines, which need to be well loaded to prevent carbon build-up reducing
performance.
Vessels with variable pitch thrusters can depend on a guaranteed base load from each
thruster of around 20% but this is not the case with variable speed drives. The solution for
vessels employing variable speed drives is to use the thrusters in bias mode (fixed azimuth
with opposing thrust vectors to create the desired resultant force) and apply significant
amounts of force bias to increase the load on the generators by having the thrusters work
against each other. This method works well and has advantages of improved station keeping
stability in benign environmental conditions and reduced wear and tear on thruster steering
gear. However, the need to manage this force bias correctly following a power plant failure
was not fully understood in some early applications. In particular, it was not always properly
controlled by the power management system, particularly if the PMS was a standalone unit
not supplied by the DP control system provider. Two issues associated with early
implementations were the need to shed the bias load before initiating overall thrust reduction
as a means of blackout prevention and also the need to shed bias in such a way that the
desired thrust vector is maintained, otherwise a drive off will result.
If the sum of the base load provided by the variable pitch thrusters and the hotel load is
larger than the rating of the largest generator on the vessel, the power plant is relatively
immune to failure to excess fuel generator faults. In this type of failure one faulty generator
takes the entire load and others trip on reverse power leading to cascade failure and
blackout. With fixed-pitch, variable speed propellers there may be times when the total
system load falls within the rating of one generator leaving the system vulnerable to this type
of failure.
5.2.7 Effect of Propeller Law and Power Factor on Post-Failure Capability
Because the relationship between propeller thrust and power is a not a straight line, as
shown in Figure 88, a vessel which is holding position with all thrusters available may need
significantly more power to hold station in the same conditions following a failure that leads
to loss of some thrusters. Thruster tripping was (and still is) a popular last resort load-
shedding feature on vessels with variable pitch thrusters. The poor low load power factor of
large asynchronous motors means that more generators have to be online even at relatively
low load, thus tripping thrusters reduces the total current demand even if the power demand
increases. On vessels with fixed pitch thrusters, which tend to have a high power factor
throughout their operating range, the advantage lies in keeping as many thrusters running as

IMCA M 206 113
possible following a failure, as it is more efficient to divide the available power between them
than to have a few thrusters working hard. In reality the advantage may be quite small and
each case needs to be considered on its merits, taking into account the power savings
associated with thruster auxiliaries which can also be tripped.
5.2.8 Regenerated Power
Other issues associated with the use of large power electronic variable speed drives are the
need to manage power regenerated by braking action. This is not usually an issue for DP but
more for transit and vessel manoeuvring when much higher levels of power can be returned
to the power plant. Some types of drives are not capable of regenerating power. With this
type of drive, care must be taken when using speed control not to reduce the speed
command set point at a rate faster than the propeller will naturally decelerate, otherwise the
inverter part of the drive will attempt to return power. Because power cannot be
transmitted beyond the drive to the power plant, the drive will only succeed in storing the
energy within itself to the point where it will be tripped by its own over voltage protection.
Where drives are designed to return power to the power generation system or to dynamic
breaking resisters, care must be taken to manage the return of this power in such a way that
generators are not tripped on reverse power or braking resistors overloaded.
This problem is often eliminated when variable speed drives are designed for true torque
control rather than speed control.
5.2.9 Effect of Harmonics
Power system harmonics have already been discussed in relation to phase shifting
transformers in 4.1.10. All variable speed drives produce harmonics of one form or another.
Generally, the more sophisticated the drive, the smaller the levels of harmonic distortion
produced. However, this is another area where the choice of thruster type can influence the
redundancy concept. If it is necessary to add harmonic filters to deal with the effects of
harmonics then it is necessary to consider the effect of these on DP redundancy.
It is notoriously difficult to specify a passive harmonic filter that will be effective in all power
plant configurations. The additional capacitance these filters add to the system may also
affect the power factor to the point where there are restrictions on the number of
generators that can be run with certain combinations of filters. If harmonic filters are to be
part of a DP redundancy concept then a very careful study of their failure effects needs to be
carried out to ensure they do not create undesirable operational restrictions.
More recently there is a trend to use variable speed drives with so called active front ends.
These are generally advertised as a solution to the problems of harmonic distortion
associated with older six- and twelve-pulse drives. These modern drives make use of
individual filters at the thruster rather than attempting to correct the entire power system.
However, even with these modern devices there have been concerns about system
resonance. It is also necessary to consider whether failure of the filter will lead to equipment
malfunction elsewhere in the plant; studies backed up by suitable testing should be carried
out to establish this.
5.2.10 Starting Transients and Inrush Current
Before solid state power electronic frequency converters of large power rating were
available, DC drives were the most popular method of obtaining speed control of motor
loads. Since this time, power electronics have advanced to the point where frequency
converters of very high power rating are available. Two device types dominate the market,
the gate turn off thyristor (GTO) and the insulated gate bipolar transistor (IGBT). The
advantages of fixed pitch propellers using variable speed drives are higher efficiency under all
operating conditions (typically in excess of 90% at full load), mechanical simplicity, improved
control and negligible starting transients. The reduction of starting transients is a significant
advantage as many motor failures can be linked to excessive heating and the large
electromagnetic forces generated by starting currents. Many motor manufacturers impose
limits on the number of starts per hour to control the thermal effects associated with direct

114 IMCA M 206
on line starting of large motors. It should be noted that some classification societies specify
that there should be no restriction on thruster starting intervals.
When variable speed drives were first introduced, one of their advantages was the ability to
soft start large motors. Prior to the application of power electronic technology, this was
achieved through the use of reduced voltage starting techniques such as star-delta and
Korndorffer starters, both of which have a significant degree of mechanical complexity.
It is fair to say that variable speed drives do remove the large starting current transient of
thruster motors when the thruster is started. This effect is achieved by ramping the speed
order up from zero to the desired speed. What is not so obvious is that there is still a large
inrush current associated with connecting the drives phase shifting or isolation transformer.
The problem has therefore been shifted from normal starting and stopping of the thruster to
blackout recovery when it is arguably more important to be able to start thrusters with as
few generators as possible. Because of the transformer inrush current it is still possible to
trip a generator if too many thrusters are started too early. An obvious solution is to add
heavy consumer start blocking, but this delays the point at which thrust can be applied until
sufficient generators are online to connect the drive transformers safely. Pre-magnetising of
transformers has been effectively used to remove inrush transients. Pre-magnetizing can be
carried out by using a small transformer to back feed the main drive transformer or by using
the drive itself powered from a large UPS or battery bank.
5.2.11 Thruster Restart
Other factors to be considered when specifying a thruster driven by a variable speed drive
are whether or not it can start the motor while it is still turning or decelerating, sometimes
called restart on the fly, and also whether it can start a thruster that is not stationary
because of inflow from sea current or adjacent thrusters. If the drive trips prior to restart it
will have to go through a pre-charge routine which can take up to ten seconds to make the
drive available again, provided it is programmed to restart automatically. All these issues can
be successfully addressed in modern drive systems but it may be more cost effective to
enquire about this feature at the time the specification is being prepared than after the
thruster is installed.
Large modern variable speed drives now offer sophisticated control systems which can
include a customer application element such that they can be used to start and stop their
own auxiliary systems such as pumps and cooling fans. This can be usefully employed to give
a thruster drive a great deal of autonomy to make itself ready for operation in receipt of a
simple request to start the thruster. Care must be taken to ensure that such control systems
are protected by a UPS with sufficient duration to keep the drive controls available during
blackout recovery. It is also worth ensuring that the ride through of auxiliary systems, such
as cooling pumps, matches that of the drive. In the worst case there is little point having a
sophisticated drive designed to ride through a voltage dip if it is tripped on loss of cooling
water flow because the contactor for its cooling water pump dropped out due to the same
voltage dip.
5.2.12 Diesel Driven Thrusters
Even thrusters which are directly driven by diesel engines are not without issues that can
influence the redundancy concept. Fixed pitch variable speed units often have clutches to
disconnect the propeller from the engine for starting. These clutches can be operated by an
engine driven hydraulic pump but can also be operated from an electric pump. Some designs
require pressure to keep the clutch engaged and should have their own accumulator to allow
for auto changeover of the pump to a standby unit. However, there have been designs where
this has not been the case and many thrusters dropped out of DP simply because one
switchboard supplying several pumps blacked out. The standby pumps powered from the
other side of the switchboard started immediately, but too late to keep the thruster from
dropping out of DP and going to idle. In such designs, what should have been a cast iron
redundancy concept employing direct diesel driven thrusters has been compromised by failing
to understand the importance of maintaining clutch pressure during auto-changeover of pump
supply.

IMCA M 206 115
5.2.13 Variable Pitch Thrusters (Controllable Pitch Propellers)
Variable pitch thrusters also have features that will influence the redundancy concept
including:
The pitch must be brought back to zero before the unit can be started;
Many designs will trip on loss of hydraulic pressure requiring a suitable delay on tripping
during standby pump changeovers etc.;
Large motors will require some form of soft starting as already discussed star-delta units,
Korndorffer starters or even power electronic soft starts are possible.
A failure mode that is sometimes overlooked is what happens to the power system if one of
the large motors starts direct and goes online without going through the star-delta sequence.
This is known to have happened in at least one case causing a very severe voltage dip which
affected the performance of other parts of the DP system leading to drift off.
The standard arrangement for fixed speed thrusters uses an induction motor to drive a
controllable pitch propeller (CPP). Selectable speed motors are also available which will
normally have two separate windings, one for high speed and one for low speed. Speed
selection in this type of machine is achieved by altering the number of poles.
Traditional methods for large high voltage machines include autotransformer and Korndorffer
starting. The high starting currents also require a substantial spinning reserve of reactive
power to restart propulsion motors. There has been at least one incident where motors
could not be restarted after a partial blackout because of the lack of spinning reserve, in spite
of the fact that there was adequate power to maintain the online thrusters at full power. The
problem was made worse by the vessel operating in open bustie configuration.
5.3 Thruster Auxiliary Systems
5.3.1 General
Thrusters need various support services including:
hydraulic power packs for blade pitch control;
hydraulic power packs for steering of azimuth thrusters;
hydraulic power pumps for retraction;
lubricating pumps;
FW and de-ionised water cooling units for motors, transformers and drives;
ventilation;
UPSs for control, pre-charge and ride through.
There are several strategies for providing power for thruster auxiliary systems. Arguably the
most secure is to closely associate the power supply for the auxiliary systems with the supply
to the thruster main motor itself. However, many vessels were, and continue to be, designed
with thruster auxiliary system powered from the main LV distribution. Where a thruster has
a duty and standby pump for a particular service, one will be supplied from the port LV
switchboard and the other from the starboard LV switchboard. This is not necessarily the
best arrangement however as it introduces vulnerability to standby start failure and certain
restriction on pump utilisation and maintenance. If the reason for providing dual pumps is to
mitigate the low reliability of the pump and not its source of supply then there is no
significant disadvantage to powering both the duty and the standby pump from the same LV
switchboard.
5.3.2 Hydraulic Power Packs
Hydraulic power packs typically consist of two or three electrically driven pumps in a duty
standby arrangement. Where the thruster requires more than one pump for normal
operation the action of pitch or steering control can become slower if operation is reduced
to one pump. This should be considered in the design of any protective functions based on

116 IMCA M 206
steering or pitch control speed, particularly if a single distribution fault can fail one pump on
each thruster. As with any essential consumer, voltage dip ride-through needs to be
considered. Figure 93 shows the schematic of a typical hydraulic steering system for a large
azimuthing thruster.
5.3.3 Thruster Prediction Errors
Most DP control systems are arranged to provide a prediction error if the thrust magnitude
or direction is not as expected. Prediction errors may occur if hydraulic systems are not
capable of turning the thruster or changing blade pitch in the expected time.

Figure 93 Thruster hydraulic system
5.3.4 Thruster Lubrication
Figure 94 shows a typical thruster lubrication system for a large azimuthing thruster. Many
thrusters of this type can operate for an extended period without forced lubrication. Some
designs of thruster impose a maximum load under these conditions. Where such restrictions
are applied automatically, the DP control system must be advised that the thruster is no
longer capable of its full capacity.
Many types of thruster depend upon compressed air to provide positive pressure on shaft
seals to prevent seawater ingress and contamination of the oil. Compressed air systems are
often common to all forward or all aft thrusters such that failure of the service air system
leads to a loss of seal pressure on all forward or all aft thrusters.

IMCA M 206 117

Figure 94 Thruster lubrication system
5.3.5 Thruster Cooling
Many thrusters require some form of fresh water cooling for things such as:
phase shifting transformers;
motors;
variable speed drives;
HPUs;
lub oil coolers;
HVAC.
In the case of high voltage variable speed drives, there may be a need for an intermediate
cooling system using de-ionised freshwater. Figure 95 shows a typical thruster FW cooling
system for a single thruster. In general, thrusters should be designed to be as independent of
each other as practical even if the redundancy concept accepts that more than one thruster
can fail due to loss of a main switchboard. Some classification societies require independent
cooling systems all equipment intended to provide redundancy and independent cooling
system for thrusters are considered to be good practice in view of the risk of losing multiple
thrusters due to leaks or maintenance related activities. If there are concerns about the
ability to carry out maintenance on freshwater cooling systems then each thruster should be
provided with a second pump and cooler rather than create commonality.

118 IMCA M 206
THRUSTER
COOLING EXP
TANK (0.5M)
THRUSTER
COOLING FW
PUMP
DRIVE
FW COOLING SYSTEM
NO.1 MAIN
TRANSFORMER
FOR THRUSTER
THRUSTER
ELECT MOTOR
AIR COOLER
THRUSTER
L.O. HEAT
EXCHANGER UNIT
THRUSTER
STEERING HPP
UNIT
NO.2 MAIN
TRANSFORMER
FOR THRUSTER
THRUSTER
CENTRAL
FW COOLER

Figure 95 Thruster cooling water unit
5.4 Thruster Motors
5.4.1 DC Motors
DC motors come in several winding configurations. DC motors for speed control are
generally separately excited machines with separate controlled rectifiers for the field winding
and armature winding. DC motors for thrusters are generally rated at 600V and power
ratings of a few hundred kW to 1MW. For larger power outputs it is common to find
tandem motors arranged to drive a common shaft. Commutators and brushes are used to
make the main power connection to the armature winding and these introduce maintenance
issues. DP vessels are still being built using DC motors and modern DC drives.
5.4.2 Asynchronous Motors
Asynchronous motors, also known as induction motors, are by far the most common type of
motor for fixed speed and variable speed drives. This type of motor has a squirrel cage rotor
and a three phase stator winding. Large motors for fixed speed drives are generally rated at
6.6kV or 11kV and power ratings up to 3MW (typical marine application). Asynchronous
motors for use with variable speed drives may be typically rated at lower voltages such as
1270V, 2.2kV or 3.3kV and power ratings up to 5MW are not untypical in marine
applications, although much larger drives can be found in other applications. A typical motor
speed range is 0 to 750 rpm but higher speeds can be found.
5.4.3 Synchronous Motors
Synchronous motors are very similar to synchronous generators in so far as they have an AC
stator winding and a DC rotor winding. They found specialist application in the Synchrodrive
which was fitted to some types of podded drive and are still specified in some higher power
applications. They were also used as synchronous condensers in some vessels to correct low
power factor.

IMCA M 206 119
5.4.4 Insulated Bearings
Large motors used for thrusters typically require at least one insulated bearing to prevent
circulating currents damaging the bearing surfaces leading to premature failure. How such
currents arise and where they flow can be complex but there are at least two mechanisms by
which this can occur. The first is related to asymmetric flux which induces a voltage in the
rotor shaft. This will circulate a current by way of the bearings if the path is not interrupted
at one point. Motors driven by variable speed drives with high frequency switching devices
such as IGBTs may suffer from capacity coupled currents which can flow through parts of the
motor normally considered to be insulators. Manufacturers may use a number of methods to
protect the bearings and care must be taken not to defeat these measures by shorting out
insulated bearings with earthing straps from instrumentation, such as shaft encoders for
example.
5.5 Variable Speed Drives
5.5.1 General
Several types of power electronic variable speed drive are available for propulsion drives.
By far the most popular is the multi level pulse width modulation (PWM) inverter but other
types of drive are specified for some applications.
5.5.2 DC Drives
DC drives are the original variable speed drive. The DC drive usually consists of a silicon
controlled rectifier driving a separately excited DC motor. They have the advantages of low
cost, simplicity and compact size. Disadvantages include a more complex motor and limited
power rating. There are still many vessels with DC drives in service and designers continue
to specify DC propulsion in some vessel projects. When choosing to specify DC drives, it is
important to note that they may be vulnerable to power system faults that cause misfire and
commutation failure, leading to loss of all thrusters if supplied from a common power system.
Some types of drive will shut down and lockout a restart attempt until a local reset is applied,
which has obvious implications for blackout recovery.
5.5.3 Cycloconverter Drive
The cycloconverter drives offer very high torque at low speed; typical applications include
icebreakers and large ships. Disadvantages include poor speed range. Figure 96 shows the
power component layout of a basic cycloconverter drive. Note that in practice, there are
several different forms of cycloconverter.

INPUT
TRANSFORMER

POSITIVE
CONVERTER
GROUP

NEGATIVE
CONVERTER
GROUP
M

Figure 96 Cycloconverter drive

120 IMCA M 206
5.5.4 Synchroconverter Drive
The synchroconverter drive is a current-source load-commutated inverter (LCI) connected
to a synchronous motor. This type of inverter is available to power ratings in excess of
50MW. The synchroconverter drive has the advantage of being electrically and mechanically
simple. Characteristics include high starting torque, good voltage dip ride through and wide
speed range. Figure 97 gives the basic power component layout of a LCI driver.

INPUT
TRANSFORMER


M
SYNCHRONOUS
MOTOR
SMOOTHING INDUCTOR
THYRISTOR CONVERTERS
FORCED COMMUTATION FOR STARTING,
NATURAL COMMUTATION WHEN RUNNING

Figure 97 Synchroconverter drive
Note that good voltage dip ride through depends on the drive controls being well protected
by UPS.
5.5.5 Voltage Source PWM Drives
Multilevel pulse width modulation (PWM) drives are now the industry standard for most
electric propulsion applications. Figure 98 shows the basic power components of a PWM
drive. PWM was developed to improve upon the current harmonics of six and twelve pulse
fully controlled bridge output stages. Although the drive output voltage of simple PWM
schemes is still a square wave, the mark space ratio is altered to simulate the effective area
under an equivalent sinusoid. The effect is a near sinusoidal motor line current waveform
with considerably fewer low order harmonics. Higher order harmonics may be increased but
these are more easily filtered.
Lower harmonics in motor line current means smooth, quiet operation and a reduction in
unwanted heating effects. Although the line current waveform of PWM drives is a major
improvement the voltage waveform is still essentially a square wave. More advanced PWM
drives use multiple step levels combined with pulse width modulation to improve the voltage
waveform. When these drive output voltage waveforms are filtered, the result is a near
sinusoidal voltage and current waveform.
Major drive manufacturers claim an overall efficiency of the order of 96% including the output
filter. PWM drives offer many other advantages such as near constant power factor
throughout the operating range; values in the region of 0.9 are typical. Many drives also offer
sophisticated motor control algorithms, some of which use mathematical models of the
motor. From a system protection standpoint, drive manufacturers offer short circuit proof
drive output converters, which means that a thruster failure is handled at the drive itself and
the upstream protection need not operate for this type of fault. Earth fault, thermal, over
current and over voltage may also be offered as standards.
Wear and tear on circuit breakers may be reduced as motor starting and stopping is handled
by the drive. Reduced arcing can also be expected.
All power electronic drives create harmonics on the system to which they are connected.
Generally speaking, the higher the order of harmonics, the more easily they are dealt with

IMCA M 206 121
and much may depend on the type of input stage specified. Six-pulse input rectifiers offer low
cost with a penalty in terms of harmonic performance. Twelve-pulse rectifiers, supplied by
drive transformers with star and delta secondary windings, are a standard way of improving
upon this. Even better performance can be obtained by increasing the pulse number yet
further, however, a cost penalty has to be accepted. Several manufacturers also offer drives
with an active front end as another way of reducing supply side harmonics.


DC LINK
OUTPUT POWER STAGE
GTO OR IGBT
OPTIONAL
OUTPUT FILTER
DELTA DELTA
WYE DRIVE
TRANSFORMER
12 PULSE DIODE BRIDGE. OR
THYRISTOR BRIDGE FOR
REGENERATIVE APPLICATIONS
M
SYNCHRONOUS
MOTOR

Figure 98 Voltage source PWM drive
5.5.6 Ride Through Performance
Although power electronic variable speed drives have been in use in DP vessels for more
than ten years, the significance of some of their features and flaws is only now being
understood by the DP community often as the result of investigations into DP incidents.
In many cases these features were well understood by the drive manufacturers but for some
unknown reason the significance of these flaws and features was not communicated to the
designers of the DP redundancy concept. One such feature is the ability (or lack of ability) of
the variable speed drive to ride through a power system transient caused by the effect of
clearing a fault elsewhere in the distribution system. This issue is also of great importance in
the process and chemical industries where plant operators do not want critical parts of the
process to trip every time there is dip in grid voltage.
Figure 99 shows a much simplified schematic of a variable speed AC thruster drive. Voltage
source drives such as this will trip on severe voltage dips to protect themselves from the
inrush current that follows power restoration. In recent years, drive manufacturers have
addressed this issue by using the power of fast control systems to stop the drive consuming
power during the voltage dip thus preventing its own internal voltage falling to dangerously
low levels. However, these features are often not tested in practice and therefore the first
real test is usually when the feature is called upon to operate in service. Had vessel owners
and designers been more aware of these flaws and features, the arguments surrounding
operation with busties open or closed may have been very different.
<U<
M
CAPACITOR
BANK
INVERTER
PROTECTION
RECTIFIER
PHASE
SHIFTING
TRANSFORMER
DC LINK
PRECHARGE
SUPPLY
HV
LV

Figure 99 Variable speed thruster drive

122 IMCA M 206
Some designers back up the voltage dip ride through capability of the drive by providing
automatic reconnection and restart. If the drive detects a significant power system
disturbance (over/under voltage, over/under frequency) it will be disconnected but will
continue to monitor the power supply until it determines that it is safe to automatically
reconnect.
5.6 Thruster Control Systems
5.6.1 Thruster Manufacturers Control Systems
Most thruster manufacturers provide a closed loop control system for their thrusters. For
azimuthing thrusters with controllable pitch propellers the closed loop control system will
accept both steering and pitch commands from the DP control system and operate the
hydraulic control system to achieve the desired blade pitch and steering angles. In the case of
azimuthing thrusters with fixed pitch propellers, the closed loop control of propeller speed is
provided within the variable speed drive. Only the closed loop steering control is provided
by the thruster control unit.
Figure 101 shows a typical thruster control unit for a fixed pitch azimuthing thruster.
The unit accepts steering commands from the DP control system in the form of an analogue
4-20mA loop representing the required angle. The control unit then interprets the command
and operates the clockwise (CW) and counter clockwise (CCW) solenoid valves to bring the
thruster azimuth to the required angle by the shortest possible route. The hydraulic system
is typically capable of turning the thruster at 2 rpm.
There are three angle measuring devices mounted on the thrusters. One device provides
feedback to the control unit for closed loop control, the other provides feedback direct to
the DP control system and the third is used for indication at the manual thruster control
levers. The closed loop feedback device may be a shaft encored in some applications.
Feedback to the DP control system is usually provided by a sine/cosine potentiometer driven
from an 10V supply as shown in Figure 100. This device has resistance elements which
produce a sinusoidal distribution of voltage. Wipers displaced by 90 allow a cosine voltage
to be produced. The angle of the thruster is then computed from :
if V_sin > 0 and V_cos > 0 angle = arctan(V_sin / V_cos) / (2 x ) x 360;
if V_cos = 0 and V_sin > 0 angle = 90;
if V_cos < 0 angle = 180 + arctan(V_sin / V_cos) / (2 x ) x 360;
if V_cos = 0 and V_sin < 0 angle = 270;
if V_cos > 0 and V_sin < 0 angle = 360 + arctan(V_sin / V_cos) / (2 x ) x 360.
Where V_sin and V_cos are the voltage signals from the sine output and cosine output in
Figure 100.
In other devices a 4-20mA interface is used to indicate the sine and cosine feedback. Loop
monitoring is provided to indicate that one channel has failed. Some control system suppliers
also carry out a check that sin
2
A+cos
2
A = 1.

IMCA M 206 123

POWER
SUPPLY
EARTH
-Ve
SUPPLY
+Ve
SUPPLY
INSTRUMENT
EARTH
SINE OUTPUT
COSINE OUTPUT

Figure 100 Sine/cosine potentiometer
Although the DP control system only provides open loop control of the thruster (closed by
vessel position) it does monitor the thruster speed and azimuth and will issue a prediction
error if either variable deviates from the required value by more than a defined amount in a
specified time. Note that at least one type of thruster has a single mechanical drive for all
three angle indicators. If this drive slips, the thruster may be pointing in the wrong direction
with no indication that this is the case.
The thruster control unit provides one of several inputs to the DP control signal. This signal
indicates to the DP system that the thruster is ready for DP commands. Other inputs may
include the variable speed drive indicating that it is ready for speed or torque commands.
Some thruster failures will cause the DP ready signal to indicate not ready. As soon as the
DP control system detects the change in status it will automatically deselect the thruster and
reduce the command to zero.
The thruster control unit will also provide some of the interlocks and protection associated
with the thruster. Typical interfaces for this purpose include:
shaft brake applied;
air pressure available;
HPU pressure;
main and backup power supply present;
lub oil pressure;
thruster control unit healthy warning/fault;
DP ready;
local/remote;
main motor start allowed;
main motor running/stopped;
stop main motor.

124 IMCA M 206
BRAKE

CW

CCW

LOCAL / REMOTE
DP READY
WARNING
FAULT
SHUTDOWN MOTOR
PROP SPEED FEEDBACK
AZIMUTH FEEDBACK TO
THRUSTER CONTROL UNIT
AZIMUTH FEEDBACK TO DP

SINE / COSINE
START ALLOWED
UPS POWER IN

THRUSTER
CONTROL UNIT
AZI CMD 4-20mA
MOTOR STOPPED

PROP SPEED
LUBE OIL PRESSURE
HPU PRESSURE
SHAFT LOCK OPEN
SHAFT BRAKE OPEN
DRIVE
VESSEL
AUTOMATION
SYSTEMS
THRUSTER
BACKUP DC SUPPLY

HYDRAULIC PUMP RUNNING
(RETRACTABLE
AZIMUTH THRUSTER ONLY)

Figure 101 Thruster control unit
5.6.2 Direct Control by Vessel Automation System
In some applications there is no thruster manufacturers control unit and the hydraulics are
interfaced directly to a vessel management system field station. The thruster control
algorithms for steering and pitch control reside within the field station. This is a popular
solution for vessel upgrades where the thruster mechanical part is to be retained but the
obsolete control system is absorbed into a new vessel automation system.
5.6.3 Thruster Emergency Stops Line Monitoring
Classification societies normally require that remote thruster emergency stops are located at
the main DP station. Emergency stops for safety purposes may be located at other points in
the thruster or drive machinery space. The thruster emergency stop needs to be
independent of the normal drive control system. Ideally the emergency stop would act
directly on the drive main circuit breaker but very few variable speed manufacturers adopt
this because shutting down the drive in this way carries a significant risk of damage.
Therefore most emergency stop functions act on the drive control system in some way,
usually as an input to the safety shutdown chain part of the drive controller electronics.
An alternative scheme has been proposed in which the drive will be shut down gracefully by
the initial action of the E stop, with a time delay circuit to open the drive circuit breaker
directly if the drive controller fails to open it in a few seconds.
DP rules and guidelines require that thrusters fail safe but there will always be some
circumstances where the DPO needs to shut down a thruster quickly using a control other
than the normal stop function. In the case of a run-away thruster it may be difficult to tell
which thruster is faulty as all thrusters may load up to oppose the faulty one. Prediction
alarms or other alarms indicting thruster faults may help to identify the faulty unit.

IMCA M 206 125
Classification society requirements vary but for DP Class 2 it is not unusual to require that all
propulsion related emergency stops use normally open contacts to prevent spurious loss of a
thruster. At least one classification society requires line monitoring to prevent shutdown of
the thruster on emergency stop cable faults such as open circuit or short circuit. To achieve
this, isolated switch amplifiers are used in combination with stop buttons having the necessary
series and parallel resistors, as shown in Figure 102. The amplifiers will only respond to the
correct change in line current caused by closing the stop button across the parallel resistor.
Alarms are provided to indicate a cable fault or loss of the emergency stop power supply.
LOCAL
24Vdc
0Vdc
E-STOP
REMOTE (BRIDGE)
PWR PWR
THRUSTER
STOP
CCT
STOP
FAULT E-STOP FAULT
CABLE
FAULT ALARM
PSU ALARM

Figure 102 Thruster emergency stops
In DP Class 3 vessels, the thruster emergency stops may form a common point connecting all
thrusters, which should be taken into account when the effects of fire are considered. Line
monitoring is generally accepted as mitigation of this potential failure. Other methods have
been accepted as reducing the risk of thrusters responding to emergency stop cable faults.
One possible alternative is to arrange the emergency stop with two control circuits one using
normally open contacts and the other using normally closed. The thruster will only shut
down if both circuit change state, alarm will be given if the two circuits ever indicate the same
status. This arrangement is similar to the logic used in the DP control systems fire back-up
switch for DP Class 3 vessels.
To provide more information on the nature of a fault, the alarms for emergency stop cable
faults should reset automatically at the switch amplifier if the fault clears. However, the
vessel management system will retain the alarm until acknowledged so that the fault can be
investigated.

126 IMCA M 206
6 Safety Systems
6.1 General Principles of Safety Systems
6.1.1 Purpose
Safety systems are designed and installed to protect personnel, environment and assets from
the consequences of abnormal and hazardous situations, and to allow a safe evacuation of
personnel within a reasonable time frame. The hazards that are considered include fire and
the presence of inflammable gas from drilling operation, onboard hydrocarbon processing or
nearby installations.
The safety systems are designed to allow for the shutdown of main power plant, thrusters,
ventilation systems and fuel and lub oil pumps. For fire incidents there may also be systems
to trigger injection of fire suppression mediums. When fire suppression systems are used
these will require ventilation shutdowns for the associated spaces.
6.1.2 Application
For installations covered by MODU and vessels such as FPSOs a shutdown system will be
required to make safe the processing plant. This system may be integrated with a common
ESD or may be a separate system.
The ESD system is used to provide a safe and rapid shutdown of systems and equipment.
The ESD system processes input signals from manual pushbuttons and selected F&G signals.
When designing these systems associated with a DP system the design must consider possible
impact of the safety systems on the DP system, including the effects of the failure modes or
acts of mal-operation.
The designer of the safety systems may prioritise the reliable operation of shutdowns to
ensure faults do not prevent a shutdown being carried out. The designers of the DP system
are more interested in ensuring that faults in the safety systems do not compromise station
keeping integrity. This section is mainly concerned with resolving that apparent conflict.
6.2 Regulations Relating to Shutdown Systems
Regulations for shutdown systems are almost entirely written from the point of view of ensuring
safety rather than minimising the risks of faults in safety systems affecting the DP system. However,
there is some recognition that the risk of unintentional stoppages should be minimised. (See MODU
Code 6.5.3, DNV-OS-A101 Safety Principles and Arrangements, Section 5, Table C1 Safest conditions
and corresponding output circuit configuration.)
A comprehensive treatment of ESD systems and fail safe functionality is given in DNV-OS-A101
Section 5. This describes how systems should be designed so that risks of unintentional shutdown caused
by malfunction or inadvertent operation is minimised. However there is no cross reference to the sort of
redundancy requirements that are now commonplace in vessel management systems.
Classification society rules require ventilation stops outside machinery spaces to be used in the event
of a fire but make no mention, directly, to the consequences of any failure in the stop system.
As discussed earlier, the DP system will be affected by failures of the auxiliary systems. For DP Class
2 that will include any failure in the shutdown systems and for DP Class 3 systems it will include the
effect of fire on associated cabling in any compartment. Rules also require the consideration of any
reasonable act of mal-operation. These requirements should mean that the shutdown and safety
systems are treated in the same way as any other part of the vessels control system, meet the same
redundancy requirements and have their failure modes analysed in the DP system FMEA.

IMCA M 206 127
6.3 ESD Systems and DP Redundancy
There are numerous ways to arrange emergency shutdown systems and fire and gas detection. Some
of these are illustrated in the diagrams below.

PLC or Relay
Cabinet
Fire and
Gas Alarm
Mimic
Diagram
Gas Detection
Emergency Stop
Panel with
Group Stops on
Mimic Diagram
Fixed Fire Suppression System Operating
Cabinets
Outputs to Ventilation Fans, Ventilation
Flaps, Fuel Pumps, Lub Oil Pumps etc.
Push
Button
Signals
Fire Detection

Figure 103 Simple centralise emergency stop system
In Figure 103 fire and gas detection is displayed on a mimic diagram above another mimic with group
and individual stop buttons. The stops are implemented via a central PLC cabinet. The main issues
here for effects on the DP system are failure effects due to hardware or power failures in the PLC
cabinet. Although the PLC may be designed so that no shutdowns occur when power fails it would be
preferable to use separate PLCs with redundant power sources split to match the redundancy
concept.
The same issues, of the system not being split to match the redundancy concept also apply to the
more sophisticated system shown in Figure 104. Here, although the F&G system uses four field
stations, the ESD system only uses two. If such an arrangement is used on a system with a four-way
redundancy split the ESD system might cause the failure of half the systems which may be worse than
the WCFDI. To make this arrangement acceptable, the system needs to be designed to fail safe.
Arranging circuits such that they do not trip on power failure greatly reduces the risk. Class may
require that this is demonstrated at FMEA proving trials for DP vessels.

128 IMCA M 206

Fire and Gas System
based on Four VMS Field
Stations
Fire Detection
Fire
Detectors
Manual
Call Points
Gas
Detectors
Push
Buttons
Fire Pumps
Fire
Fighting
Systems
ESD and F&G Matrixes
Incorporating ESD 1
and ESD 0 buttons.
Located in Bridge and
ECR
ESD System Based on
two VMS Field Stations
VMS System Based on Field
Stations and Redundant
Network for Control of Power
System, Auxiliaries and DP.
Uses separate Field stations for
each Thruster and Each
Switchboard.
Redundant Data/
Network Links
ESD Level 0 Pushbuttons at
Lifeboat Muster Stations
Fire Dampers
Ventilation Fans
Valves
Equipment
Shut Down
ESD 0 Abandon Platform Shutdown
ESD 1 Preparation to Abandon
ESD 2 Local ESD F&G Pushbuttons
Local Emergency Stop Buttons in
Machinery Spaces (ESD 2)

Figure 104 ESD and F&G system integrated into VMS
An ESD 0, as shown in Figure 104, causes a total shutdown of all systems. Some would consider such
an arrangement an unacceptable risk and have a policy to not implement ESD 0.
Measures that may be included in such a system to reduce risk include:
additional subdivision of hardware to operate all stop functions along the lines of the redundancy
concept;
use of two buttons to avoid accidental operation;
use of covered type buttons;
use of an enable keyswitch which can be in the off position during DP operations;
ESD set to manual activation only during DP operations;
field station outputs fail safe on power loss to prevent tripping of DP critical items.
An arrangement split along the lines of the redundancy concept (in this case a four-way split) is shown
in Figure 105. Here any fault in the ESD should only affect one of the four switchboards.
The ESD system shown in Figure 105 is divided into four, along the lines of the redundancy concept.
This means that the worst failure that is likely to occur would be a spurious shutdown of 25% of the
power and thruster systems. There is a small risk that a severe event affecting one of the lifeboat
stations could affect all engine rooms. This risk could be minimised by operating the system inhibits
on the bridge during DP operations. Providing alarm and delay on ESD 0 to allow the control room
operators or DPOs to cancel the shutdown has also been implemented in some applications.
Note that the system is provided with a separate disable facility to each controller on the bridge
(bridge enable control). This is a recently introduced safety measure required to enhance security.
The difficulties with the design of ESD 0 when applied to DP drilling units arise because the rules were
written for moored units and thus station keeping was of little concern. A moored unit will not drift
off on loss of power but blackout of a DP MODU (because of a gas cloud) means that the vessel drifts
downwind remaining in the gas cloud.

IMCA M 206 129
In the design of an ESD system there may be a risk of a blackout due to a wiring fault. These risks can
be minimised by using:
line monitoring;
dual circuits with NO and NC contacts which must change state together;
voting systems;
ensuring that trips do not occur if power is lost to the controllers or field stations.
10 c
ESD HVA
ENABLE
ESD BVB
ENABLE
ESD HVC
ENABLE
ESD HVD
ENABLE
Bridge Enable Control
LQ Vent Shutdown
ESD-LQ-PB
Bridge Control Panel
MS Vent Shutdown
A
ESD-MS-PBA
MS Vent Shutdown
C
ESD-MS-PBC
MS Vent Shutdown
B
ESD-MS-PBB
Main DP/Back-UP
DP Alt. Vent S/D
ESD-PB-DP
11 kV Swbd
MS Vent Shutdown
D
ESD-MS-PBD
E-GEN Room Vent
Shutdown
VESD-EGEN-PBD
480 SWBD RM A
11KV SWGR RM A
VMUPS1
10 c
11 kV Swbd
480 SWBD RM C
11KV SWGR RM C
VMUPS3
10 c
11 kV Swbd
480 SWBD RM D
11KV SWGR RM D
VMUPSD
10 c
11 kV Swbd
480 SWBD RM B
11KV SWGR RM B
VMUPS2
24V
PSU
FS08
24V
PSU
FS07
24V
PSU
FS06
24V
PSU
FS05
Water Mist Valve Operation Panel in ECR
Stbd Pump Rooms, Engine Rooms 1 4,
Thruster Rms 1, 3, 5 and 7
To CPA
To CPA, CPB, CPC and CPD
To CPA
To CPB
To CPC
To CPD
To CPD
To CPA
To CPB
To CPC
To CPD
ESD HVA
PB-2
ESD BVB
PB-2
ESD HVC
PB-2
ESD HVD
PB-2
AFT Lifeboat Station
To CPA
To CPB
To CPC
To CPD
Trip Signals to 440V ac
Consumers
Trip Signals to 440V ac
Consumers
Trip Signals to 440V ac
Consumers
Trip Signals to 440V ac
Consumers
ESD HVA
Controller
CPA
ESD HVB
Controller
CPB
ESD HVC
Controller
CPC
ESD HVD
Controller
CPD
To CPD
To CPA
To CPB
To CPC
To CPD
ESD HVA
PB-2
ESD HVC
PB-2
ESD HVD
PB-2
FWD Lifeboat Station
To CPA,
CPB,
CPC and
CPD.
VESD-DF-
PB1
Drill Cabin
Water Mist Valve Operation Panel in ECR
Stbd Pump Rooms, Engine Rooms 1 and 2,
Thruster Rms 1, 3, 5 and 7
Water Mist Valve Operation Panel in ECR
Port Pump Rooms, Engine Rooms 1 4,
Thruster Rms 2, 4, 6 and 8
Water Mist Valve Operation Panel in ECR
Port Pump Rooms, Engine Rooms 1 4,
Thruster Rms 2, 4, 6 and 8
To CPB
To CPC
To CPD
ESD HVB
PB-2

Figure 105 ESD 0 split to match redundancy concept
6.4 Active Fire Protection
6.4.1 Design Considerations
Active fire suppression systems include fire suppression gases and water mist systems. These
are usually activated from individual control panels mounted immediately outside the space
covered. These panels are in turn linked to some arrangement that will shut off the
ventilation and close ventilation flaps (fire dampers) related to the area containing the fire.
With this design, any fault in these panels should only affect the ventilation for one space.
However, if a central relay panel is used, the failure effects from that panel may affect more
than one space if not carefully designed. When designing these systems it is important to

130 IMCA M 206
consider the effect on DP related systems caused by intentional operation and by system
failure.
For DP Class 2 vessels it is generally accepted that failure of the firefighting system must not
lead to a failure effect exceeding the worst case failure design intent even if intentional
operation will have that effect. However, it is good practice to split firefighting systems and
their controls along the lines of the redundancy concept. This is particularly valid in the case
of DP Class 2 vessels with more than one engine room. DP Class 3 vessels are designed to
withstand the effects of fire in any one compartment and firefighting systems should be
designed in a manner that supports the redundancy concept.
6.4.2 Typical System
Figure 106 shows a typical CO
2
system for a large DP Class 2 vessel with two engine rooms.
The operating philosophy outlined below describes some of the features used to help ensure
the correct space is flooded and prevent unintentional actions such as:
stop of fans;
closing of dampers;
CO
2
discharge.
In this example, the starboard engine room requires 44 CO
2
cylinders and the port generator
engine room requires 47 CO
2
cylinders. One control valve cabinet is allocated to each
engine room and each cabinet houses two manually operated valves. One of the two valves
valve directs gas from the pilot bottles to activate all the cylinders downstream of the CO
2
control line through a pneumatic timer. The other valve directs pilot gas to keep the main
stop valve in the open position. A limit switch on the main stop valve will indicate to the
F&G system that the valve has been opened.
The pilot cylinders are located in the cylinder control cabinet, accessed by a key. On a
confirmed fire in a CO
2
protected space, a red light will be illuminated on the corresponding
control valve cabinet to direct the operator to the correct cabinet. Ventilation shutdown and
alarms are initiated by limit switches on the cabinet doors.
Opening the control valve cabinets for the port or starboard engine rooms is detected by
two limit switches to reduce the risk of spurious operation, this triggers the following actions:
a CO
2
release column light alarm will be activate in the associated space;
the bridge will receive audible and visual alarms on the F&G system.
Opening the cylinder control cabinet activates a third limit switch which initiates:
CO
2
release alert on the F&G system;
trip of ventilation fans and closure of fire dampers (provided all three limit switches are
activated).
More advanced features can be included such as linking initiation of the CO
2
release process
to the power management system which will start standby generators in the unaffected
engine room and open the bustie when the process is complete, thus reducing the risk of a
fire affecting both power systems.

IMCA M 206 131

Figure 106 Typical CO
2
firefighting systems for large vessel
6.5 Effects of Ventilation System Shutdown
6.5.1 Machinery Spaces
In spaces such as thruster rooms and auxiliary machinery spaces there should not be an
immediate effect on systems due to a spurious shutdown of ventilation but this should be
confirmed during FMEA proving trials.
Measures to mitigate the longer term effects can include alarms for unscheduled fan stops,
temperature monitoring of machinery and room temperature sensors. Trials should check
that the measures used give a reliable and timely warning of ventilation failure well before
machinery must be shut down to avoid damage.
6.5.2 Engine Rooms
See Section 2.3.6 for a discussion of engine room ventilation shutdown. Restriction of air
flow into the engine room can have a more serious effect on engine operation and can also
represent a safety hazard in relation to slamming of doors etc.
Figure 107 shows a typical fire damper system for a large DP Class 2 vessel with two engine
rooms. The fire dampers are pneumatically controlled using two independent sources of air.
The dampers for each space are controlled by dedicated F&G process stations. Due to the
importance of maintaining combustion air to the engines, the fire dampers for the engine
rooms are arranged to fail to the open position on loss of air supply or control signal. The
fire dampers for other spaces fail to the closed position.

132 IMCA M 206
ECR
Fire Dampers
S
S
S
S
S
S
S
S
S
S
S
S
S
From Aft QCV/RCV
Control Air System
Port Gen
Eng.Rm Fire
Dampers
S
S
S
HV/LV Swbd
Room (Port)
Fire Dampers
S
S
S
S
S
S
S
S
S
S
From Aft QCV/RCV
Control Air System
Stbd Gen
Eng.Rm Fire
Dampers
S
S
S
HV/LV Swbd
Room (Stbd)
Fire Dampers
No.2 Aft
Fire
Damper
Air
Reserv.
No.1 Aft
Fire
Damper
Air
Reserv.
Reduc
er
F&G Process Station
F&G Process Station F&G Process Station
F&G Process Station
Fail
Open
Fail
Open
Fail
Close
Fail
Close
Fail
Close
S
S
Fail
Close
Aft Machinery
Space Port
S
S
Fail
Close
Aft Machinery
Space Stbd
F&G Process Station F&G Process Station

Figure 107 Fire damper controls
6.6 Group Stops
Group stops are used to simplify operation of multiple functions to shut down equipment for an area.
These should be split along the lines of the redundancy concept with respect to both to the initiating
buttons and the hardware (cables, cable routes, relays, PLCs, field stations etc.) used to act on the
stop command. The type of shutdown circuit used may be determined by class requirements including
the provision of line monitoring.
6.7 Control Room Layouts
Modern DP and vessel management systems have tended to locate the main operator stations on the
bridge, it is therefore sensible to install the matrix panels and the controls for emergency shutdowns,
fire and gas detection etc. at the same location. This allows senior personnel to control and monitor
all aspects of an ongoing incident from one position. The ergonomics of this arrangement are
important to reduce the risks of shutting down the wrong systems. These facilities will usually be
duplicated in the engine control room (ECR) where such a location exists.
The arrangement of the VMS control may be set such that acknowledging engineering alarms is the
responsibility of personnel in the ECR and all safety related (for example, fire and gas alarms), DP and
ballast control alarms are to be the responsibility of the bridge, with two DPOs manning the DP and
VMS desks.

IMCA M 206 133
Forward DP and Navigation Desk
FWD
Main DP Desk Main VMS Desk
E
S
D

M
a
t
r
i
x

a
n
d

S
a
f
e
t
y

P
a
n
e
l
s

Figure 108 Positioning of ESD/safety systems
In the example in Figure 108, the ESD matrix and safety system has been positioned so that senior
personnel in control of an emergency situation on the bridge can best control the situation. Also,
routine false alarms, zone isolations etc. can be dealt with and monitored by the DP operators easily.
VMS system operations dealing with F&G alarms, hot work isolations, ballast control etc. can be a full-
time task on a large rig therefore positioning the VMS desks near to the F&G/ESD panels makes this
task easier.

134 IMCA M 206
7 Vessel Management Systems
7.1 General Description
7.1.1 Distributed Control System
The vessel management system will manage the functions of control, monitoring and alarm
management of vital machinery required to control the vessel including engine and propulsion
auxiliary systems, fluid and cargo systems and other ancillary systems. A modern vessel
management system will utilise a distributed control system to perform these functions.
As the name implies a distributed control system does not concentrate control of all
processes at a central location, rather it distributes the control of tasks to different nodes
(PLCs or RCUs). These nodes will normally be housed close to the plant they are
controlling, thereby cutting down on wiring. Suitable communication networks will link all
aspects of the system together.
DCS architecture generally consists of one or more HMI, PLCs or RCUs, I/O modules and a
data communication highway linking the PLCs to the HMI and vice versa. The I/O modules
within the controller nodes will use appropriate communication for the plant they are
controlling. This could be analogue or digital I/O as required. Figure 109 shows a basic DCS
with A60 compartment separation to allow some continued operation following the loss of
any compartment to fire or flood etc.
RCU or PLC
and IO
modules
RCU or PLC
and IO
modules
RCU or PLC
and IO
modules
RCU or PLC
and IO
modules
RCU or PLC
and IO
modules
Operator
Console 1
Operator
Console 2
Operator
Console 3
RCU or PLC
and IO
modules
Dual Ethernet Network
Sensor
Sensor
Pump
Valve
Sensor
Sensor
Pump
Breaker
Sensor
Sensor
Relay
Valve
Sensor
Sensor
Pump
Valve
Pump
Relay
Sensor
Sensor
A60 Separation

Figure 109 Basic distributed control system
In reality it can be argued that current systems are a SCADA/DCS hybrid, in as much as the
software within the HMIs will monitor aspects of the plant, alarm when an event is detected
and carry out some supervisory action such as starting a standby pump.
This is the basis of a SCADA system where data gathering is the prime objective and actions
are event driven. A DCS is usually accepted as being process driven, meaning that all tasks
allocated to a node are normally run sequentially, usually within a ladder programme of a
PLC, and mismatch alarms etc. are not generated until the process looking at that particular
I/O is run.
7.1.2 Industrial Networks
Industrial networks is a general term given to any type of communication connecting devices
for process control applications. Industrial networks can be divided into several types:
control networks;
fieldbus networks;
sensor networks.

IMCA M 206 135
Control network: Sometimes called the process network, this is the high level network
used to transmit instructions and data between the HMI and the PLCs which control the
plant. Different technologies and protocols have been used over the years but the preferred
solution used by most companies at this time is Ethernet/IP, where the IP stands for industrial
protocol.
It has been argued that Ethernet/IP is a non-deterministic protocol and when the system is
using network hubs, half duplex, CSMA/CD architecture this is true to some extent,
depending on overall bandwidth usage. However with newer installations, where data
switches are used in place of hubs, and full duplex 100M networks are in use, the network
can be considered deterministic for all practical purposes. The use of switches also allows
additional quality of standard (QoS) rules to be implemented, i.e. the switches can assign
higher priorities to specific traffic.
Fieldbus network: The origin of the fieldbus was to replace hard wired connections
between complicated field devices and their controllers by a single digital link where all data
could be transmitted in a serial string. To accommodate information from several sources
the data is time multiplexed.
P
S
U
C
P
U
A
n
lo
g
u
e

I
n
p
u
t
D
ig
it
a
l
I
n
p
u
t
A
n
lo
g
u
e

O
u
t
p
u
t
D
ig
it
a
l
I
n
p
u
t
START / STOP / RESET
ALARM DATA
RUNNING
VOLTAGE / CURRENT /FREQUENCY
BAUD RATE
PLC / RTU MOTOR CONTROL CENTER
PROFIBUS
SERIAL LINK

Figure 110 Fieldbus communications using Profibus DP protocol
With all devices or equipment that support the Profibus DP protocol, a general system data
(GSD) file (text file) should be supplied by the manufacturer. This GSD file contains general
information about the device, type of device, supported baud rates, a summary of functions
supported within the protocol and all cyclic data (data that is exchanged every bus scan with
the master).
Figure 110 shows how this operates in practice. The PLC will have the MCC as one of the
devices it must interrogate and control during its scan. Data recovered from the MCC will
be repackaged and sent to the HMI on the control network.
Sensor network: This is the most basic network and works by detecting the status of a
sensor and transmitting this information as a discrete 1 or a 0 to the PLC. This might be the
position of a valve (open or closed) or the state of switch.
7.1.3 Network Topologies
With different types of networks as described above there are different network topologies.
Historically, these were designated as star, bus and ring, however as technology has
progressed these have changed somewhat from their original design and a more correct
name for each one is now:
physical star logical bus;
physical bus logical ring;
physical ring logical ring.
Physical star logical bus topology: In its simplest form a star network consists of a single
device connected to each node by a separate cable, as shown in Figure 111. In earlier
fieldbus iterations this would have been all field devices possibly connecting to a single main

136 IMCA M 206
frame computer in the control room. All control functions would be sent to the field devices
using point-to-point hardwired connections. This was a true star topology as all data would
flow back through the mainframe. The system was expensive due to the amount of wiring
involved, complex to configure and difficult to maintain. In addition the lack of effective
standards led to expensive upgrades when new technology was introduced.
5 6 3 2 1 4
Main Frame
Computer
HMI
7 8 11 10 9 12
Field
Devices

Figure 111 Star topology
With the introduction of distributed control and agreements on standards for industrial data
communications new models were required. Figure 112 shows a small network with three
switches and twelve nodes. Although each switch may physically look as though it is the
centre of a star, internally each switch has a linear bus as a backbone, making this a physical
star logical bus topology.
Switch 1 Switch 3 Switch 2
12
11
5
6
8 9
3 2
10
1 4
7

Figure 112 Physical star logical bus
Some of the main advantages of this network are scalability with no disruptions to the
network when adding or removing devices and easy fault detection. In addition, the failure of
a node will have no effect on the rest of the network. With the use of switches and 100M
full duplex communication a node communicates with any other node using point-to-point
communication. Each node believes it has total use of the bandwidth, so no collisions occur
and retransmitting of data is not required.
In addition to contributing to the advantages of this topology, the switches can be seen as the
main disadvantage. Failure of a switch will mean any nodes connected to it will be
unreachable. As discussed above each node believes it has total use of the bandwidth,
therefore although the bandwidth used between nodes in point-to-point communication may
be low, if several nodes are communicating simultaneously the cabling between the switches
may become a bottleneck. This is usually overcome by using fibre optic cabling between
switches connected to Gigabit uplink modules within the switches. As modern switches
normally have a Gigabit backbone this effectively means the network can accommodate
10 x 100M individual networks.
The disadvantages discussed above are overcome by utilising a dual network, nominally called
net A and net B, Figure 113 illustrates this configuration. The two networks are completely
independent with separate switches, separate cabling and dual isolated network adapters
within the network nodes. Separately sourced power supplies are also provided for switches
on the different networks. All data is transmitted on both networks simultaneously with the

IMCA M 206 137
same time stamp. Only one message is processed and the other one is dropped with no
further action being carried out.
This dual redundant topology is ideal for Ethernet connections and is used extensively for
control networks, i.e. between the HMIs and the RCUs etc.
Switch 1
(Net A)
Switch 3
(Net A
Switch 3
(Net A
12
11
5
6
8 9
3 2
10
1 4
7
Switch 1
(Net B)
Switch 3
(Net B)
Switch 2
(Net B)

Figure 113 Dual redundant star topology
Physical bus logical ring topology: In this topology all nodes are connected to a shared
backbone using multi-drop lines connected to medium access units (MAU) as shown in Figure
114. The MAU devices ensure the connected node is ready for network traffic before it is
allowed on to the bus. Data placed on the bus is propagated to all operating nodes and is
therefore sometimes described as a broadcast system. A terminating load (resistor) is fitted
at each end. This is required to optimise signal quality and prevent signal reflection.
Incorrect termination can make a bus unusable especially at high transmission speeds.
5 6 3 2 1 4
11 12 9 8 7 10
MAU

Figure 114 Bus topology
This topology greatly reduces cabling costs when compared to a star network. It is easy to
implement and fault finding is relatively easy. Failure of a node should have minimal effect on
the rest of the network.
There are several disadvantages of this basic system, the main one being that as a single
backbone is used only one data packet can be transmitted at any time. All transmissions
should therefore be carefully controlled by some arbitration system. This explains why most
industrial bus systems use some type of token passing system.
A basic token passing system would consist of a data object that is passed around the nodes
in a ring. When a node receives the token it has a finite time (token hold time) to take
control of the bus, to either send data or receive data. It will send any high priority packets
first and if any time is left will send any low priority packets. Once the token hold time has
expired the next lowest addressed node receives the token and the cycle continues.
The token passing protocol is easy to implement in software, in addition when adding new
nodes to the network the ring is reinitialised and the target token rotation time (time for a
complete cycle around the bus) is recomputed.
The token passing bus is not a physical bus but a logical ring, in other words once the token
has been passed to node 12 on our diagram it is then handed to node 1, thereby completing
the ring. Figure 115 illustrates this setup.

138 IMCA M 206
Physical Bus
Logical Ring
5 6 3 2 1 4
11 12 9 8 7 10

Figure 115 Physical bus logical ring
Physical ring topology: Figure 116 shows a basic ring network where each node connects
to the nodes directly adjacent to it on both sides forming a single continuous route around
the ring. Data travels in one direction with each node regenerating the network packets and
forwarding them to the next node. A token passing system is used as discussed above,
except that the token is passed to the next physical node as opposed to logical node, as seen
in the physical bus logical ring topology. If any node is not switched on, electronics within the
MAUs ensure data is still forwarded to the next node.
5 6 3 2 1 4
11 12 9 8 7 10
MAU

Figure 116 Physical ring topology
Advantages of this arrangement are that data transmission is structured with every node
having access to the token for a specific time. Large networks can be configured without
complex mapping being required.
Historically, logical and physical ring token passing networks performed better than star
networks at heavy network loads due to the complexities and time delays introduced by the
retransmissions of data caused by the CSMA/CD aspect of a star network. However this
advantage has been negated by the use of full duplex point-to-point operations as discussed
above. Typical ring transmission rates are 4MHz or 16MHz which means they are
considerably slower than a star/bus topology using fast Ethernet which will operate at
100MHz.
Because the basic ring topology is uni-directional a failure of a single cable, multi-drop link or
node could cause severe disruption to network traffic. To overcome this problem it is now
common practice to install a counter-rotating ring to provide a redundant topology as shown
in Figure 117. Until recently this was uneconomic due to the cost (and perhaps weight) of
the additional cabling. With the introduction of cheap fibre cabling more vendors are offering
this solution as an alternative to the star/bus system, arguing that it is a more deterministic
solution.

IMCA M 206 139
5 6 3 2 1 4
11 12 9 8 7 10
RING 2
RING 1

Figure 117 Dual ring topology
7.2 Network Technology
7.2.1 General
There are many vendors offering electrical, power and control systems, and there are as
many different network technologies on offer. A vessel management system will have several
networks installed, perhaps an Ethernet control network, a Profibus DP fieldbus network and
a Modbus RTU sensor network. Gone are the turnkey solutions of the past where the
owner was tied into a particular vendor for the life of the vessel.
Standards based solutions have allowed several network protocols to come to the fore. This
section discusses several of the most popular technologies in use and examines some of the
advantages and disadvantages of each. Before discussing the differences, it is useful to
consider what each of the networks has in common that allows them to be used extensively
in an industrial environment.
Offshore industrial network systems are subject to environmental factors and other design
requirements not normally included in the design of an office network. Some of these are:
predictability;
reliability in a harsh environment;
minimum downtime;
ease of maintenance/repair.
Predictability: The system must have some degree of determinism. As systems operate in
a real time environment any failure or alarm must be reported and acted on quickly enough
to prevent any knock-on effect further affecting the system. The network topology plays a
part in this determinism. Token ring networks and star/bus networks operating in full duplex
can be considered deterministic.
Reliability in a harsh environment: Offshore environmental factors including, vibration,
heat, salt-laden atmosphere, electrical noise, etc. must be taken into account when designing
the network system.
Minimum downtime: If a network is unavailable, some systems or devices may stop
communicating. At a minimum this will mean redundancy is compromised. The network
system should have been in service long enough for any inherent design flaws to come to the
fore or to have been stress tested to ensure mean time between failures is acceptable.
Ease of maintenance/repair: A well designed system should have built-in diagnostics that
enable the electrical or instrument technicians to quickly pinpoint where system failures have
occurred. Most vendors now provide some type of net status page or mimic on the HMI to
assist fault finding. Where possible, modules should be designed to allow them to be

140 IMCA M 206
swapped out either without switching off the rest of the network, or by isolating just the
faulty section.
Other issues which may influence the choice of a particular network include cost
effectiveness, whether it complies with relevant standards, scalability and ease of use. These
are budgetary considerations and are not considered further here.
7.2.2 Introduction to Network Protocols
It is not possible to discuss network protocols without mentioning the Open Systems
Interconnection (OSI) reference model. This is a seven layer model developed in the 1970s
to attempt to standardise network communications.
Network protocol describes the functions that happen within these seven layers. The
following discussion is a very broad description of the functions of each layer of the model
with Figure 118 illustrating a typical transfer of data between two operator stations.
PHYSICAL
APPLICATION
PRESENTATION
SESSION
TRANSPORT
NETWORK
DATA LINK
OPERATOR A
PHYSICAL
APPLICATION
PRESENTATION
SESSION
TRANSPORT
NETWORK
DATA LINK
OPERATOR B
NETWORK
Data transfer
A to B
NETWORK

Figure 118 OSI network model
Physical layer: Provides the connection to the media, electrical and physical. This layer
also provides the data link layer with the means to transmit a serial bit stream on to the
media. This might be RS232, RS485 or transmissions from or to a network interface card.
Data link layer: Defines the network topology to be used, chops the data into byte-size
packets and wraps them for transmission by adding frame headers and footers and unwraps
and reassembles packets coming back in before passing them up to the network layer. The
data link layer is also responsible for the first line of error checking and retransmission of
missed packets.
Network layer: Primarily used for routing, it takes the data from the layer above (transport
layer) and adds the network address of the destination, passes it down to the data link layer
and removes superfluous information on packets received from the data link layer before
passing up the chain.
Transport layer: Responsible for reliable end to end user communications. Typical
functions include flow control (ensuring a reliable connection has been established between
devices) and multiplexing or slotting data from several applications into a single packet.
On receipt of the packets from the network layer, it removes any remaining bits relating to
data communications and formats the messages into a language which is not protocol
dependent.
Session layer: Establishes, maintains and terminates network sessions by allocating system
resources as required. Will provide error reporting for the Application and Presentation
Layers.
Presentation layer: If required, translates the incoming data sent to the application layer
into a format it will understand. For example, if the data is a text message to be displayed on
the screen the presentation layer may ensure the output is sent as an ASCII character. If it is
a diagram, it may send as a GIF or a JPG.

IMCA M 206 141
Application layer: This is basically the user interface, if the command sent from the
operator was to start a pump this is the instruction sent to the controller, RCU or PLC.
It is worth noting at this point that most industrial networks do not implement all seven
layers and actually use a three-layer model. Layer 3 (network) is normally not implemented
as no routing to external networks is required. Layer 4 (transport) is not needed as the data
link layer (level 2) carries out similar packet sequencing functions, which is adequate for the
complexity of these networks. Level 5 (session) is not required as high level software (e.g.
token passing system) can be used to control session lengths etc. Level 6 (presentation) layer
is not required as all devices and nodes will be communicating in the same language.
7.2.3 Ethernet/IP
Ethernet/IP is the preferred option at this time for the control network; Ethernet/IP
implements layer 1, 2, 3 and 7 of the OSI model. Current systems utilise a full duplex
industrial Ethernet network operating at 100MHz. The protocol used is TCP/IP.
The purpose of the Ethernet/IP network is to transmit control data between the units
comprising the vessel management system. All operator stations (HMI) and field stations are
referred to as data terminating equipment (DTE). Each DTE has one or more network
interface adapter. Each node has a unique static internet protocol address allocated when
the interface card is installed. The use of IP addressing requires the inclusion of layer 3
(network) in this model.
Operation: Data on an Ethernet network is transmitted across the medium in serial using a
non-return-to-zero signal. The encoding scheme is differential Manchester code. The
concepts and description of how Manchester code operates is beyond the scope of this
document.
The network interface card acts as a transceiver. The transceivers within each node provide
the following:
a medium access control (MAC) unit, which is responsible for frame wrapping/
unwrapping and error detection (OSI layer 2);
a RAM module to allow the medium access control unit to talk to both the network
cable and the host computer at a high bit rate (OSI layer 2);
electrical isolation between the cabling and the interface electronics (OSI layer 1);
protection of the network from malfunctions in the transceiver (OSI layer 1).
This last function is also referred to as jabber control since, without the appropriate
protection electronics, if a fault develops, a faulty transceiver may continuously transmit
random data (jabber) onto the network and inhibit or corrupt all other transmissions. This
has great significance for DP control systems and most system providers now implement
measures to prevent this.
Network collisions are avoided by utilising fast Ethernet in full duplex mode for network data
transmissions. Fast Ethernet utilises two pairs of wires from CAT 7 shielded twisted pair
(STP) cabling. In older systems with half duplex mode, one pair of wires is used to send or
receive data, and the other pair is used to listen on the network for other nodes transmitting
(collision detection). In full duplex mode one pair of wires is used to transmit and one to
receive.
The use of network switches ensures the network packet is only sent to the correct address.
When a switch is first connected to the network it builds up a routing table. When messages
are received on a particular port the from address is noted in the routing table; the switch
will initially forward this message to all ports except the port the message was received on.
Only the destination port will acknowledge receipt of the message so the switch can further
build up the routing table. Eventually all ports will have entries in the routing table and the
data through the switch will be effectively point to point.
CRC and other validation checks are carried out. These include checking the frame is neither
too short nor too long. If any of these checks fails the frame is discarded and an error status

142 IMCA M 206
sent to a higher level and eventually to some type of net status page. Assuming no frame
errors, the destination address field is processed.
Some data is still required to be sent to several or all units. This can be done by sending a
broadcast packet. During normal operations broadcast packets will be forwarded to all
ports, however if the number of packets exceeds a preset threshold in a preset time, packets
will be discarded for a set time. When this time has passed the adapter will resume accepting
packets until the time threshold is exceeded again.
All correctly addressed messages are processed at the network layer and if it is determined
the data is for the current node, the frame is passed to layer 7 for further processing.
7.2.4 Profibus
Profibus provides three different versions of the communication protocol: factory message
specification (FMS), decentralised peripherals (DP) and process automation (PA). A brief
description of each is given below, however only the operation of Profibus DP is described in
any detail. Note that the abbreviation DP is not to be confused with dynamic positioning in
this case.
Profibus FMS: In this model layers 1, 2 and 7 are implemented. FMS was developed as a
multi-peer network for communications between PLCs, HMI and field devices. This general
purpose solution could therefore be used in control networks, fieldbus networks and sensor
networks. Due to several factors, including the limited baud rate not being able to supply the
required refresh rates for modern HMI, it is not being implemented in new control networks.
In addition the newer Profibus DP uses a similar transmission medium so FMS is now mainly
used to support legacy equipment.
Profibus DP: Implemented at layers 1 and layers 2 and optimised for high speed, this
version was designed especially for fieldbus operations and is used extensively in modern
vessel systems.
Profibus PA: Profibus PA is designed for the sensor networks to connect sensors, relays,
actuators etc. to a common fieldbus. Implemented at layers 1 and layers 2, Profibus PA
transmission techniques are in accordance with standards to ensure intrinsic safety and bus
powering of field devices even in potentially explosive areas.
Operation: Profibus DP can operate in a mono-master or multi-master mode depending on
configuration. Basically one or more master (active node) controls a number of slave devices
(passive node) and polls them as required. Although the Profibus protocol includes token-
passing software Profibus DP only uses this if it is configured in multi-master mode.
The token is only passed between the masters, not the slaves. The procedure for token
passing is discussed above in bus topologies. Once an active node receives the token it
communicates with the passive nodes using a master-slave setup.
Profibus DP normally operates using a cyclic transfer of data between master and slaves on an
RS485 network. The master-slave procedure permits the master to periodically poll each
node assigned to it. All data communication between a master and slave originate from the
master device. Each slave is assigned to one master and only that master may write output
data to that slave. Profibus DP is most often set up as a single mono-master configuration,
however, in a multi-master configuration any master may read information from any slave,
but masters can only write data to their own assigned slaves.
The RS 485 specification for Profibus is based on semi-duplex, asynchronous transmission.
The transmission medium can be shielded twisted pair (STP) or fibre optic cabling. When
using STP the maximum line length is determined by transmission speed, as shown in Table 9.
Baud Rate (kbit/s) 9.6 to 187.5 500 15,000 12,000
Segment Length (m) 1,000 400 200 100
Table 9 Profibus maximum segment length

IMCA M 206 143
Data is transmitted in an 11-bit character frame in non-return to zero (NRZ) differential
Manchester code similar to Ethernet. As each slave is polled cyclically there are no collisions
and large amounts of data can be transferred in single telegrams.
Vcc Vcc
Node 1
(active)
Node 2
(passive)
Node 3
(passive)
32 Nodes/Segment maximum
3
9
0

3
9
0

3
9
0

3
9
0

2
2
0

2
2
0

RS 485 Segment
GND GND
Node 4
(passive)
Polling
1 1 1 1 0 0
B
A
0 0
Data Byte

Figure 119 RS 485 bus segment
In Figure 119, the bus line is a shielded twisted pair cable which is terminated at both ends. A
maximum of 32 nodes can be connected to any segment. Line B high/line A low indicates a
binary 1. The start bit is always a low and the stop bit and any idle bits are always a high.
Data transferred between the master and slave is formatted according to the generic station
description (GSD) file for the particular master or slave device. GSDs files must be provided
by the device vendor to meet Profibus standards; however Profibus International has also
defined GSDs for many standard devices.
Advantages of Profibus: Profibus is the most widely accepted and installed industrial
network for fieldbus applications and is supported almost universally by all vendors. It was
developed in 1989 and has a stable architecture. Platform independence means any number
of master and slave devices are available off the shelf. The polling nature of Profibus means it
is a truly deterministic network. The FMS, DP and PA versions cover all aspects of the
industrial network.
Disadvantages of Profibus: When only transferring small amounts of data there is a high
overhead and it has slightly higher installation costs than some other networks. The main
Profibus version (DP) has no power on the bus. Due to the requirements for termination
resistors the bus topology used is sensitive to wiring problems and installation problems
when upgrading. However with the increased acceptance of ProfiNet, basically an Ethernet
version of Profibus DP most if not all of these disadvantages will be overcome.
7.2.5 Modbus
Modbus is briefly discussed here as it is used by most PLC manufacturers in addition to their
own proprietary language. Modbus is an application layer (layer 7) messaging protocol.
Communication is based on a query-response protocol where there is one master and one
or more slaves. Modbus was originally created as a way for main frame computers to gather
information and control the operation of the PLCs connected to them, a basic SCADA
system.
Although not all Modbus slave devices are PLCs, they are still composed of two basic
components. These are the central processing unit (CPU) and the I/O system interface. The
CPU section is designed to carry out a programmed set of instructions in a pre-determined
order. This programme will normally be stored as a ladder diagram within a memory map
and the PLC will descend the ladder one rung at a time. Information will be sent to or
requested from the field equipment.
The I/O system is the interface between the field equipment and the controller. Incoming
signals from limit switches, relays, pumps etc. are wired to terminals of the input interfaces,
where if required they are passed through analogue to digital converters, and the resultant
data is stored in 16-bit unsigned (positive integer) addressable registers until the CPU via the
ladder program requests the data.

144 IMCA M 206
16-bit unsigned addressable registers are also used to store the information for devices to be
controlled, when the PLC ladder programme instructs the control signals to be output, the
data is connected to the terminals of the output modules, this time through a digital to
analogue converter if required.
The Modbus protocol was designed to provide a way to transfer the contents of these
registers to another PLC or a host PC. Serial Modbus connections use two basic protocols,
these are Modbus ASCII and Modbus RTU. Both methods use layer 1, 2 and 7 of the OSI
model.
Modbus ASCII: When using Modbus ASCII, all messages are coded in hexadecimal values,
represented with readable ASCII characters. For every byte of information, two
communication-bytes are needed, because every communication-byte can only define four
bits in the hexadecimal. As discussed above, the protocol was designed to pass 16-bit
register values, so it can be represented as four hexadecimal numbers each made up of four
bits of data.
The data is then sent for framing using a process called ASCII encoding. A colon (:) character
begins the message frame and a carriage return/line feed signifies the end of the message.
A longitudinal redundancy check (LRC) and a parity bit provides assurance that the data is
not corrupted in transmission.
Modbus RTU: The RTU model was developed to provide faster transfer of data. When
using Modbus ASCII the messages are encoded in hexadecimal. When using RTU the data is
exchanged in binary, making the transfer more efficient. In addition better error checking is
carried out with RTU as CRC checking is carried out instead of LRC. Table 10 lists the
properties of the two protocols.
Modbus ASCII Modbus RTU
Characters ASCII 0-9 and A-F Binary 0...255
Error Checking LRC longitudinal redundancy CRC cyclic redundancy
Start Frame : (colon) >28 bits idle time
End Frame CR/LF >28 bits idle time
Fragment Error 1 second / no data 12 bits / no data
Start Bit 1 1
Data 7 8
Parity even/odd none even/odd none
Stop Bit 1 2 1 2
Medium RS232 / RS422 / RS485 RS232 / RS422 / RS485
Table 10 Modbus ASCII and RTU properties
Modbus RTU messages are framed differently to ASCII messages. Each message is preceded
and ended by a time gap >28 bits. If a receiver detects a gap of at least twelve bits during a
transmission this is assumed to be an error and the receive buffer is cleared in preparation
for a new message.
Although both protocols use the same medium (RS232/RS422/RS485), it is not possible to
mix Modbus ASCII and Modbus RTU on the same wire. It should also be noted that although
both ASCII and RTU support RS232 and RS422, if these mediums are used only point to
point communication can be established, i.e. only one slave device can be connected.
To allow multiple slaves RS 485 is required.
Modbus also has a TCP/IP version which is essentially binary data (RTU) packaged within the
TCP/IP protocol. This allows for faster data over greater distances; however it is not used to
any extent in offshore systems at this time.

IMCA M 206 145
The main advantages of Modbus are its easy deployment, and simple configuration, allowing
legacy equipment to be easily interfaced with newer technologies. Its universal acceptance by
vendors in all regions of the globe means Modbus RTU along with Profibus DP and
Ethernet/IP are seen as the de facto standards for industrial networks.
7.2.6 WorldFIP
WorldFIP is the protocol from which the original Fieldbus standards were developed.
The protocol is based on a producer/distributor/consumer (PDC) model with distributed
arbitration, rather than the master/slave or token passing systems discussed earlier.
The WorldFIP protocol like the majority of the other protocols is a three-layer model
operating at layers 1, 2 and 7 of the OSI model.
The WorldFIP physical layer ensures the transfer of information on the bus.
The transmission medium is either shielded twisted pair or optical fibre. The network
topology is a physical bus, which means all devices are attached to the same line. Data is
transmitted using differential Manchester code and transmission speed is 31.25Kbps, 1Mbps,
2.5Mbps and 5Mbps (fibre only).
There are two types of node in the WorldFIP network. These are the bus arbitrating nodes
and the producer/consumer nodes. The bus arbitration (BA) node, controls accessibility to
the bus and signals to the other nodes when they can communicate and is the distributor in
this model. At any instant in time a subscriber can be either a producer or a consumer.
If a producer, it is placing data on the network to be picked up or consumed by other
subscribers on the network. If a consumer, the controller listens on the networks and
receives the data.
For redundancy purposes it is common practice to configure several nodes to perform the
task of bus arbiter, however only one BA can be active at any time. Normally the bus arbiter
is selected arbitrarily the first controller capable of being a bus arbiter that ascertains that
the network has no bus arbiter will assume the role. Once the system is operational and a
BA has been identified it commences a bus scan. A scan involves the constant repetition of
the FIP macrocycle. The macrocycle contains a scanning table, a list of identifiers to scan and
the periodicity associated with each identifier.
The BA runs through the scanning table sequentially. It broadcasts the name of the identifier
on the network. One and only one PLC recognises itself as the producer of the identifier.
One or more PLCs recognise they are consumers of the variable to be transmitted.
The producer then broadcasts the value of the identifier and all consuming stations
simultaneously capture the variable.
Bus
Arbiter
Producer /
Consumer
Producer /
Consumer
Producer /
Consumer
Producer /
Consumer
Identifier 1 = Node 01 = Producer
Consumers = Node 03 and Node 04
Node 01 Node 02 Node 03 Node 04
Bus

Figure 120 FIP network configuration
Figure 120 describes the BA putting identifier 1 on the bus. Node 01 recognises this as a
variable within his databanks and prepares to put this on the bus as the producer. Nodes 3

146 IMCA M 206
and 4 identify themselves as consumers and ready themselves to accept the variable. The BA
waits the correct amount of time to allow the transactions to occur and then places the next
identifier on the network. This scanning table is read through and then the operation is
repeated.
This type of transmission is classed as periodic transfer, in addition to this, at preconfigured
time slots within the macrocycle, time may be allocated to allow subscribers to transfer
aperiodic data, making this both a process driven and event driven protocol.
It appears from recent designs that the WorldFIP protocol is being used less and less in
offshore installations, but is considered here due to the legacy equipment still in use.
7.2.7 CANbus
Another producer/consumer network that is gaining popularity, particularly in engine
monitoring systems and with some control system vendors, is CANbus (controller area
network). CANbus uses layer 1 and layer 2 of the OSI model and was originally developed in
the late 1980s to simplify the design of wiring harnesses within automobiles by moving from
point-to-point connections to using a multi-drop bus topology. Since then several industrial
fieldbuses have been developed using CAN as their underlying technology, these include
DeviceNet, ControlNet, CANopen and CAN Kingdom. The fieldbuses mentioned are not
compatible with each other and the following description is of the messaging protocol within
the basic CAN network.
Unlike most other protocols, in basic CAN the data packets do not contain either a
transmitting or receiving node address. Instead each message contains a unique identifier
based on the content of the message and its priority. This identifier is assigned during system
design. Parameters might be RPM, temperature, frequency etc. It is expected that some
parameters will change more rapidly than others and therefore need to be transmitted more
frequently, these will be allocated the higher priority. The identifiers allocated the lowest
physical values (binary) have the highest priority.
A node with data to transmit will place it onto the bus, it is possible that more than one node
will attempt to transmit simultaneously, therefore some type of arbitration is required. The
method used is a version of CSMA/CD similar to half duplex Ethernet, except that where
Ethernet uses destructive bus arbitration, CANbus uses non-destructive bitwise arbitration.
With Ethernet, on detection of collision all parties retreat from the bus for a random amount
of time and retransmit, hoping the bus will be free. With CANbus any bus conflicts are
resolved with the identifier with the lowest value transmitting, and all other nodes halting
transmission. A wired-and mechanism (collector dotting) is used to determine the message
with the lowest identifier, where the dominant state (logic 0) overwrites the recessive state
(logic 1); nodes not transmitting, or with a recessive transmission, automatically become
receivers of the transmission. The message is transmitted as a broadcast with all receivers
interrogating the message and using the data as required.
The drawing at Figure 121 illustrates the wired-and non-destructive arbitration in operation.
Nodes A-D are transmitting simultaneously, arbitration is necessary and the signals are
wired-anded. The first node to pull back and cease transmitting is node C, followed by
node B then node D. Node A continues transmitting and the bus output reflects this.
One advantage of this system is that bus allocation is determined by need and negotiated only
between the messages attempting to transmit. As there is no requirement for silence on the
bus as with Ethernet CSMA/CD, or the time scheduling required with token passing, this non-
destructive bitwise arbitration usually ensures no bandwidth is utilised without the
transmission of useful data.

IMCA M 206 147
A
D
C
B
Stops Transmitting
Stops Transmitting
Stops Transmitting
Bus
Recessive
Dominant
Vcc
pull up
resistor
Node A
(010101101)
Node B
(010110110)
Node C
(011110111)
Node D
(10101101)
CANbus (Wired-And)

Figure 121 CANbus arbitration
CANbus has been in use for over fifteen years and a perceived disadvantage was the limited
data transfer rate, however with the number of fieldbuses developing this technology it is
now possible to achieve data transfer up to 5 Mbps. As all identifiers are allocated a physical
number during the design phase, it may be expensive to scale the system up during mid-life
upgrades.
7.3 Redundancy
7.3.1 Introduction
As control systems have become more complex the number of installed devices has
multiplied, with each device potentially affecting the overall reliability of the system. System
failures, whether due to hardware or software problems may cause downtime or
compromise safety. To minimise downtime, vendors have developed products and solutions
to provide fault tolerant systems by the use of redundant hardware or software.
Hardware solutions include the use of redundant power supplies, redundant processors,
redundant I/O modules, multiple HMIs and redundant cabling. Software solutions might
include using separate servers to install supposedly identical software or using different
versions of software in master and slave controllers or PLCs. The figure below illustrates a
typically distributed field station with redundancy built in, and the discussion below describes
the different ways redundancy can be achieved. Figure 122 shows a typical arrangement.
Power supply redundancy: In Figure 122 each PSU is fed from a different UPS. These
UPS would normally be configured in a redundant fashion with each one being supplied from
different sides of the bus. Each PSU is feeding separate PLCs and I/O racks providing
complete isolation of supply. One disadvantage of this setup is that the failure of a PSU will
cause loss of a PLC and an I/O rack. An alternative power distribution configuration would
be to feed the output of both power supplies to each module using decoupling diodes, as
shown in Figure 123 to prevent a short circuit in one PSU affecting the other. The diodes
should be tested periodically to ensure a hidden failure does not compromise redundancy.

148 IMCA M 206
P
S
U C
P
U
P
S
U
C
P
U
P
S
U
A
N
A
L
O
G
U
E

I
N
P
U
T
D
I
G
I
T
A
L

I
N
P
U
T
D
I
G
I
T
A
L

O
U
T
P
U
T
A
N
A
L
O
G
U
E

O
U
T
P
U
T
P
S
U
A
N
A
L
O
G
U
E

I
N
P
U
T
D
I
G
I
T
A
L

I
N
P
U
T
D
I
G
I
T
A
L

O
U
T
P
U
T
A
N
A
L
O
G
U
E

O
U
T
P
U
T
NET A
NET A
NET B
NET B
PSU 1
PSU 2
24V DC
(PSU 1)
24V DC
(PSU 1)
24V DC
(PSU 2)
24V DC
(PSU 2) to/from Field IO
120V UPS 1
120V UPS 2
PLC 2
PLC 1
IO Rack 1
IO Rack 2
Field Station
Terminal Board
IO Rack 1 IO Rack 2
PLC 1 PLC 2
Redundant Field Station
C
O
M
M
U
N
I
C
A
T
I
O
N

M
O
D
U
L
E
C
O
M
M
U
N
I
C
A
T
I
O
N

M
O
D
U
L
E

Figure 122 Typical redundant field station
PSU 2
PSU 1
PLC 1
PLC 2
IO Rack 1
IO Rack 2
PLC 1
PLC 2
IO Rack 1
IO Rack 2
+
+
-
-
24V
0V

Figure 123 Redundant power supplies
PLC redundancy: This can be designed in several ways. The most common way is in a
master/slave or hot standby arrangement where status is continuously monitored and data is
synchronised between both controllers. On a failure of the master there will be a bumpless
transfer to the slave. Only one controller actually outputs to the I/O modules.
An alternative way is to allow both processors to act as peers and, with both controllers
outputting data to the I/O modules, voting is carried out using software algorithms and one
or other of the signals is used.
I/O module redundancy: Identical racks can be configured to ensure each line of field I/O
is duplicated; in this way the total failure of a single rack will not affect the operation.
Although this may give excellent fault tolerance it is expensive and somewhat complicated to
implement. A more usual configuration is to split the I/O logically between the racks or I/O
modules. Loss of a rack or power supply to a rack may have an effect on the redundancy of
the system, and some field I/O will be lost, but the correct mapping of the I/O will ensure
fault tolerance is maintained. For example, from Figure 122, if pumps are operating in a
duty/standby configuration, control of one pump will be by I/O rack 1 and the other pump by
I/O rack 2. Although loss of I/O rack 1 will cause loss of communication with the duty pump,
the system can still start the standby pump as a precaution if required. Fail as set is the
preferred failure mode for most propulsion related equipment such as pumps and cooling
water valves. Some classification societies require a pulse-to-start, pulse-to-stop control
strategy.

IMCA M 206 149
Sensor redundancy: Some critical systems may have twin sensors located close together,
or even as part of the same unit. When the same action is carried out, if either or both
sensors register a signal the sensors can be considered part of the redundancy system. Once
again it enhances redundancy if the outputs from the sensors are mapped to different I/O
modules.
Data communication redundancy: Data communication redundancy was discussed in
the section on network topologies. The favoured solution at this time for the control
network is Ethernet in a star/bus topology. It is normally installed as a dual independent
system nominally net A and net B. All aspects of the network are duplicated including cabling,
switches, Ethernet adapters, network interface cards (NIC). The drawing at Figure 124
illustrates this by showing separate net A and net B cabling to each of the PLCs. Within the
communication module on the PLC there will be individual Ethernet interface adapters.
Figure 124 shows a typical DP/VMS control network. The network switches and any medium
converters (STP to fibre) are housed in network distribution units (NDU). It can be seen
that losing any single node on the network or any active component (like a switch) will not
affect the operation of the overall system as communication is still operational on the
alternative network.
NDU-A2
NDU-A1
Aux FS
FS 43
FS 44
PMS FS
FS 41
FS 42
DP Control
DPC
DP OS1
DP OS2
VMS OS
OS31
OS32
History Station
Thruster FS
FS 31
FS 32
FS 33
NET B
VMS OS
OS34
OS35
Thruster FS
FS 34
FS 35
FS 36
NDU-B1
NDU-B2
NET A
NDU Network Distribution Unit
VMS Vessel Management System
OS Operator Station
Thruster FS Thruster Field Station
Aux FS Auxiliary Field Station
PMS FS Power Management Field Station
DPC Dynamic Positioning Controller

Figure 124 Typical control network
Normally there is no redundant cabling to the field I/O or even between field stations and
main units like the generator control panels or MCCs. Communication is normally via a
single Modbus or Profibus connection. However, on safety critical systems such as fire and
gas, where a fieldbus connects the vendor specific equipment to the VMS system for
activation of CO
2
, closing of dampers and ventilation etc., it is normal to have a dual Profibus
link for redundancy purposes.
7.3.2 Alarm and Monitoring
Most DCS vendors provide an alarm and monitoring system as an integral part of their
delivery, if not it will be available as an optional extra. The main purpose of the alarm and
monitoring system is to give the operators the basic alarm and status information they
require to maintain safe and efficient operation of the plant. Information relating to power
management, propulsion, ballast control, HVAC, safety systems etc. should all be available.
To provide this data the distributed control system processes information from a multitude
of different sources. It is not unusual for a system to interrogate over 2000 separate I/O
devices and large vessels may have upwards of 5000 I/O.
Alarms: Built-in diagnostics should ensure that inconsistencies in expected results will be
detected and reported. These inconsistencies may be due to faulty field equipment, faulty
wiring, logic errors, incorrect configuration etc. The operator is made aware of these
anomalies by the use of audio and visual alarms. The audio alarms are normally buzzers at

150 IMCA M 206
the VMS operator stations (OS). This is usually a generic alarm that the operator will silence
locally at one of the HMI. The audio alarm is accompanied by a visual alarm on a reserved
part of the screen at the operator station.
As all operator stations are peers, the visual alarm will show at each station. This is normally
a banner alarm with a brief description of the fault and the tag number or I/O module
generating the fault. Although different vendors have different systems, the alarm is normally
colour coded with separate colours for severity of fault (yellow or red). Safety critical faults
may have a different coloured banner.
Alarm printers are provided to give immediate hardcopy reports on alarms and incidents.
Historically these were parallel port dot matrix printers with a continuous form feed output.
In new or upgraded systems these are being superseded by network fed single sheet feed
laser printers. This network is usually an Ethernet network connecting each HMI to the
printer. This is a separate network from net A and net B, discussed earlier in the control
network, and is normally referred to as net C or the admin net. There is no requirement
for redundancy in this network as there are no control functions involved.
Monitoring: Continuous monitoring of control functions is carried out by the alarm and
monitoring system with all alarms and process events stored in a database within each
operator station. Relevant parts of this history log can be called up within user-defined time
slices and all alarms and events displayed. The operator can then use a search string to
retrieve specific information.
To assist in fault analysis a history station can be provided, where in addition to alarms and
process events, selectable vessel management parameters are recorded for a length of time
decided by the operator. Information can then be offloaded to external media for in-depth
analysis offsite, or fed into a simulator to recreate a specific situation. Software within the
operator stations also allow real time trending to be carried out for most power
management and propulsion parameters.
1.26k
2.74k
Switch Open: i=6mA
Switch Closed: i=19mA
0V
24V
RCU 1
RCU 2
IO Module
Process Station Field IO

Figure 125 Typical line monitoring circuit
Further monitoring is carried out including line monitoring of discrete inputs. The simplified
drawing at Figure 125 shows a basic line monitoring circuit with a single field input to
redundant RCUs. The line to the switch is active at 24V. With the switch open the current
in the circuit will be 6mA, with the switch closed the current will be 19mA. Any other signal
on the line will be incorrect and raise an alarm. It should be noted this circuit is for
illustration of the principle only. In a real situation the resistor values would be different to
take the resistance of the wire and impedance matching etc. into consideration.

IMCA M 206 151
Appendix 1
Abbreviations List
A
ABS American Bureau of Shipping
AC Alternating current
ACB Air circuit breaker
ACCU Automatic control centralised unmanned
AFE Active front end
AHU Air handling unit
AHV Anchor handling vessel
AMOT Name of valve manufacturer
ANSI American National Standards Institute
ASCII American Standard code for Information Interchange
AVR Automatic voltage regulator
B
BA Bus arbiter
BTT Bow tunnel thruster
C
CA Certifying authority
CAN Controller area network
CB Circuit breaker/control breaker
CW Clockwise
CCW Counter clockwise
CD Carrier detect/collision detect
CO
2
Carbon dioxide
CoS Chamber of Shipping
CPP Controlled pitch propeller
CPU Central processing unit
CR Close relay
CRC Cyclic redundancy check
CSMA Carrier sense multiple access
CT Current transformer
D
DBR Dead bus relay
DBSR Dead bus slave relay
DC Direct current
DCS Distributed control system
DG Diesel generator
DGS Diesel generator set
DGPS Differential Global Positioning System
DI Digital input
DNV Det Norske Veritas
DO Diesel oil
DP Decentralised peripheral when used as Profibus DP
DP Dynamic positioning
DPC Dynamic positioning console/cabinet
DPO Dynamic positioning operator
DPS Dynamic positioning system

152 IMCA M 206
DTE Data terminating equipment
DTL Definite time lag
E
E0 E Zero DNV notation for unmanned machinery space
ECR Engine control room
EG Emergency generator
EGB Electric governor backup
EPD Electrical power distribution
ER Engineroom
ESD Emergency shut down
F
F&G Fire and gas
FIP Factory interface protocol
FMEA Failure mode and effect analysis
FMECA Failure modes and effects criticality analysis
FMS Factory message specification
FO Fuel oil
FS Field station
FW Fresh water
FWC Fresh water cooling
Fwd Forward
G
GIF Graphics Interchange Format
GPS Global Positioning System
GSD Generic station description
GTO Gate turn off (thyristor)
H
HF High frequency
HFO Heavy fuel oil
HMI Human machine interface
HO Heavy oil
HP High pressure
HPP Hydraulic power pack
HPR Hydro-acoustic position reference
HPU Hydraulic power unit
HT High temperature
HTFW High temperature fresh water
HV High voltage
HVAC Heating, ventilation and air conditioning
Hz Hertz
I
I> Low set current
I>> High set current
I/O Input/Output
IAS Integrated Automation System
ICMS Integrated control and monitoring system
ICS Integrated control system
IDMT Inverse definite minimum time
IEC International Electrotechnical Commission
IGBT Insulated gate bipolar transistor

IMCA M 206 153
IMCA International Marine Contractors Association
IMO International Maritime Organization
IP Internet Protocol
IP Industrial protocol
ISM International Safety Management
ISO International Standards Organisation
J
JB Junction box
JPG JPEG Joint Photographic Experts Group
JW Jacket water
K
Kbps Kilo bits per second
kN Kilo newton
kV Kilo volt
kVA Kilo volt ampere
kVAr Kilo volt ampere reactive
kW Kilowatt
L
LAL Low level alarms
LCI Load commutated inverter
LCR Inductance (L), capacitance (C), resistance (R)
LED Light emitting diode
LHS Left hand side
LO Lub oil
LOA Length over all
LR Lloyds Register
LRC Longitudinal redundancy check
LS Load sharing
LT Low temperature
LTFW Low temperature fresh water
LV Low voltage
M
mA milliAmps
MAC Medium access control
MAP Main alarm panel
MARPOL Marine Pollution (International Convention for the Prevention of Pollution From Ships,)
MAU Medium access unit
MBC Micro biological contamination
Mbps Mega bits per second
MCB Miniature circuit breaker
MCC Motor control centre
MCCB Moulded case circuit breaker
MCOS Manual changeover system
MCR Maximum continuous rating
MDO Marine diesel oil
MFR Multi function relay
MGE Main generator engine
MGP Multi generator protection
MMI Man machine interface
MODU Mobile offshore drilling unit
MRU Motion reference unit

154 IMCA M 206
MSB Main switchboard
MSC Maritime Safety Committee
MTC Manual thruster controls
MUX Multiplexer
MVA Mega volt ampere
MVAr Mega volt ampere reactive
MVR Manual voltage regulator
MW Megawatt
N
NC Normally closed
NDE Non drive end
NDU Network distribution unit
NIC Network interface connector/card
NO Normally open
NPS Negative phase sequence
NRZ Non return to zero
O
O
2
Oxygen
O/C Open circuit
OIM Offshore installation manager
OLE Object linking and embedding
OLM Optical link module
OPC Object linking and embedding for process control
OPLS Oil pressure low shutdown
OS Operator station/outstation
OSI Open system interconnection
OSV Offshore supply vessel
OT Operator terminal
P
PA Power available
PC Personal computer
PCU Process control unit
PDC Producer/ distributor/ consumer
PID Proportional integral and differential
PLC Programmable logic controller
PMG Permanent magnet generator
PMS Power management system
PS Process station
psi Pounds per square inch
PSU Power supply unit
PWM Pulse width modulation
Q
QC Quick closing
QCV Quick closing valve
QoS Quality of service
R
RAM Random access memory
RCS Remote control system
RCU Remote control unit
RHS Right hand side

IMCA M 206 155
RMS Route mean squared
ROV Remotely operated vehicle
RP Reverse power
RPM Revolutions per minute
RTD Resistance temperature device
S
s Second(s)
S/C Short circuit
SCADA Supervision control and data acquisition
SCR Silicon control rectifier
SLD Single line diagram
SMS Safety management system
Stbd Starboard
STP Shielded twisted pair
SW Sea water
SWBD Switchboard
SWG Standard wire gauge
T
TC Thruster controller
TCP/IP Transmission control protocol/internet protocol
TDAVR Thyristor divert automatic voltage regulator
THD Total harmonic distortion
TMCC Thruster motor control centre
TMS Thruster management system
TW Taut wire
U
UHF Ultra high frequency
UMS Unattended machinery space
UPS Uninterruptible power supply
V
V Volts
VAr Volt ampere reactive
VAS Vessel automation system
VCB Vacuum circuit breakers
VDU Visual display unit
VENT Ventilation
VHF Very high frequency
VFD Variable frequency drive
VMS Vessel management system
VSD Variable speed drive
VT Voltage transformer
W
WCFDI Worst case failure design intent
Z
Z Impedance

Оценить